output
stringlengths 7
3.46k
| input
stringclasses 1
value | instruction
stringlengths 129
114k
|
|---|---|---|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A machine learning system may be implemented as a set of trained models. A set of trained models, for example, a deep learning system, is disclosed wherein one or more types of media input may be analyzed to determine an associated engagement of the one or more types of media input. |
|
A method for autonomic group decision making may include presenting a plurality of decision options and receiving at least one decision option selected by each decision maker. A decision result may be presented including an indication of the preference for each option. Each decision maker may be allowed to revise the previously selected option for convergence to a single decision option for the group. An impedance score is determined for each decision maker in response to a decision maker selecting a different decision option. The impedance score may be based on one or more actions by the decision maker regarding selecting the decision option. A level of the impedance score indicates either progress toward or away from convergence. The impedance score may be revised each time a revised option is selected. The single decision option may be presented in response to the decision makers converging on the single decision option. | Please help me write a proper abstract based on the patent claims. CLAIM: 1.-10. (canceled) 11. A system for autonomic group decision making, comprising: a processor; a group decision making module operating on the processor, the group decision making module being configured to perform a set of functions comprising: presenting a plurality of decision options to each decision maker of a group of decision makers; receiving at least one decision option selected by each decision maker from the plurality of decision options; presenting a decision making result to each decision maker, the decision making result comprising an indication of the preference for each of the decision options by the group of decision makers; allowing, each decision maker to revise the at least one decision option previously selected by the decision maker for convergence to a single decision option by the group of decision makers; determining an impedance score for each decision maker in response to at least one decision maker selecting a different at least one decision option from one previously selected, the impedance score of each decision maker being based on at least one action of a set of actions by the decision maker regarding selecting the at least one decision option, wherein a level of the impedance score indicates either a progression toward convergence to the single decision option or away from convergence to the single decision option by the decision maker associated with the impedance score; revising the impedance score for a particular decision maker each time the particular decision maker selects a revised at least one decision option; and presenting the single decision option in response to the decision makers converging on the single decision option. 12. The system of claim 11, wherein a lower impedance score indicates the at least one action of the set of actions by the decision maker progressing toward convergence to the single decision option and a higher impedance score indicates the at least one action of the set actions by the decision maker moving away from convergence to the single decision option, the set of actions for lowering the impedance score comprising: changing to the at least one decision option selected by a greater number of decision makers; a shorter time duration between selecting the revised at least one decision option than selecting a previous at least one decision option in progression toward convergence of the single decision option; selecting a larger number of preferred decision options than previously selected; foregoing an option to rank preferred selections of a group of decision options; and modifying the at least one decision option to accommodate the selection of other decision makers for converging on the single decision option in less than a preset time period. 13. The system of claim 12, further comprising an impedance engine for determining and revising the impedance scores. 14. The system of claim 11, wherein the decision module is further configured to perform a set of functions comprising: presenting a revised decision making result each time at least one decision maker selects a different at least one decision option from one previously selected until the group of decision makers converge on the single decision option or expiration or a preset time period before the decision makers converge on the single decision option; determining a revised impedance score for affected decision makers in response to the at least one decision maker selecting the different at least one decision option; and presenting the revised impedance score for each affected decision maker to at least the affected decision maker. 15. The system of claim 14, further comprising selection engine, the selection engine being configured to select the single decision option in response to the expiration of the preset time period, the single decision option being selected based on the at least one decision option selected last by each decision maker and the impedance score of each decision maker, wherein the impedance score of the decision maker is used to weight the last at least one decision option by the decision maker to provide a score, the lower impedance score corresponding to higher weight, the at least one decision option with the highest score being selected as the single decision option. 16. The system of claim 14, further comprising a selection engine, wherein the at least one decision option selected by each decision maker comprises a multiplicity of decision options selected by each decision maker, and wherein the single decision option is selected in response to expiration of the preset time period the selection engine being configured to perform a set of functions comprising: determining a weighting associated with each decision maker based on the impedance score of each decision maker; assigning a value to each of the multiplicity of decision options based on a preference for a particular decision option by each decision maker, a highest preferred decision option being assigned a highest value and a lowest preferred decision option being assigned a lowest value; calculating to score for each of the multiplicity of decision options selected by each decision maker by multiplying the weighting associated with the decision maker by the value of each of the multiplicity of decision options selected by the decision maker; and adding the scores from each decision maker for each respective decision option of the multiplicity of decision options selected by each decision maker, wherein the decision option having the highest score is selected as the decision option. 17. The system of claim 11, wherein a new impedance score is dynamically determined during a decision making process based on the convergence to the single decision option in response to a different at least one decision option being selected by at least one decision maker of the group of decision makers. 18. A computer program product for autonomic group decision making, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a device to cause the device to perform a method comprising: presenting a plurality of decision options to each decision maker of a group of decision makers; receiving at least one decision option selected by each decision maker from the plurality of decision options; presenting a decision making result to each decision maker, the decision making result comprising an indication of the preference for each of the decision options by the group of decision makers; allowing each decision maker to revise the at least one decision option previously selected by the decision maker for convergence to a single decision option by the group of decision makers; determining an impedance score for each decision maker in response to at least one decision maker selecting a different at least one decision option from one previously selected, the impedance score of each decision maker being based on at least one action of a set of actions by the decision maker regarding selecting the at least one decision option, wherein a level of the impedance score indicates either a progression toward convergence to the single decision option or away from convergence to the single decision option by the decision maker associated with the impedance score; revising the impedance score for a particular decision maker each time the particular decision maker selects a revised at least one decision option; and presenting the single decision option in response to the decision makers converging on the single decision option. 19. The computer program product of claim 18, wherein the method further comprises: presenting a revised decision making result each time at least one decision maker selects a different at least one decision option from one previously selected until the group of decision makers converge on the single decision option or expiration of a preset time period before the decision makers converge on the single decision option; determining a revised impedance score for affected decision makers in response to the at least one decision maker selecting the different at least one decision option; and presenting the revised impedance score for each affected decision maker to at least the affected decision maker. 20. The computer program product of claim 18, wherein the at least one decision option selected by each decision maker comprises a multiplicity of decision options selected by each decision maker, and wherein selecting the single decision option in response to expiration of the preset time period comprises: determining a weighting associated with each decision maker based on the impedance score of each decision maker; assigning a value to each of the multiplicity of decision options based on a preference for a particular decision option by each decision maker, a highest preferred decision option being assigned a highest value and a lowest preferred decision option being assigned a lowest value; calculating a score for each of the multiplicity of decision options selected by each decision maker by multiplying the weighting associated with the decision maker by the value of each of the multiplicity of decision options selected by the decision maker; and adding the scores from each decision maker for each respective decision option of the multiplicity of decision options selected by each decision maker, wherein the decision option having the highest score is selected as the single decision option. 21. The system of claim 11, wherein the decision making module is configured to perform a further set of functions comprising: receiving a multiplicity of decision options selected by each decision maker, wherein the multiplicity of decision options are ranked by each decision maker based on a preference for each of the multiplicity of decision options; assigning a value to each of the multiplicity of decision options from each decision maker based on the preference for each of the multiplicity of decision options by each decision maker; and determining the single decision option based on the impedance score for each decision maker and the assigned value of each of the multiplicity of decision options of each decision maker. 22. The system of claim 21, wherein assigning the value to each of the multiplicity of decision options from each decision maker comprises assigning a highest preference value to a highest preferred decision option by each decision maker and assigning a lowest preference value to each lowest preferred decision option. 23. The system of claim 11, wherein allowing each decision maker to revise the at least one decision option selected by the decision maker for convergence to the single decision option comprises presenting a graphical user interface to each decision maker comprising: a number of decision makers that selected each decision option; the impedance score for at least the decision maker; and a feature for selecting a different at least one decision option from the at least one decision option previously selected by the decision maker. 24. The system of claim 16, wherein the weighting for a particular decision maker is determined by 1−(Idm−Imin/Irange), wherein Idm is the impedance score of the particular decision maker, Imin is a minimum impedance score and Irange is a range from a lowest impedance score to a highest impedance score for the group of decision makers. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method for autonomic group decision making may include presenting a plurality of decision options and receiving at least one decision option selected by each decision maker. A decision result may be presented including an indication of the preference for each option. Each decision maker may be allowed to revise the previously selected option for convergence to a single decision option for the group. An impedance score is determined for each decision maker in response to a decision maker selecting a different decision option. The impedance score may be based on one or more actions by the decision maker regarding selecting the decision option. A level of the impedance score indicates either progress toward or away from convergence. The impedance score may be revised each time a revised option is selected. The single decision option may be presented in response to the decision makers converging on the single decision option. |
|
G06N5045 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method for autonomic group decision making may include presenting a plurality of decision options and receiving at least one decision option selected by each decision maker. A decision result may be presented including an indication of the preference for each option. Each decision maker may be allowed to revise the previously selected option for convergence to a single decision option for the group. An impedance score is determined for each decision maker in response to a decision maker selecting a different decision option. The impedance score may be based on one or more actions by the decision maker regarding selecting the decision option. A level of the impedance score indicates either progress toward or away from convergence. The impedance score may be revised each time a revised option is selected. The single decision option may be presented in response to the decision makers converging on the single decision option. |
|
A cognitive information processing system environment which includes a plurality of data sources; a cognitive inference and learning system coupled to receive a data from the plurality of data sources, the cognitive inference and learning system processing the data from the plurality of data sources to provide cognitively processed insights, the cognitive inference and learning system further comprising performing a learning operation to iteratively improve the cognitively processed insights over time; and, a destination, the destination receiving the cognitively processed insights. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A cognitive information processing system environment comprising: a plurality of data sources; a cognitive inference and learning system coupled to receive a data from the plurality of data sources, the cognitive inference and learning system processing the data from the plurality of data sources to provide cognitively processed insights, the cognitive inference and learning system further comprising performing a learning operation to iteratively improve the cognitively processed insights over time; and, a destination, the destination receiving the cognitively processed insights. 2. The cognitive information processing system environment of claim 1, wherein the plurality of data sources comprise at least one of situational data sources and temporal data sources. 3. The cognitive information processing system environment of claim 1, wherein: the plurality of data sources comprise at least one of a social data source, a public data source, a private data source and a device data source. 4. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises an semantic analysis portion, the semantic analysis portion performing analysis operations to achieve a semantic level of understanding about language by relating syntactic structures. 5. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises a goal optimization portion, the goal optimization portion performing multi-criteria decision making operations to achieve a goal. 6. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises a collaborative filtering portion, the collaborative filtering portion filtering for information through collaborative involvement of a plurality of entities. 7. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises a common sense reasoning portion, the common sense reasoning portion inherently making deductions from commonly known facts. 8. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises a natural language processing portion, the natural language processing portion enabling natural language interactions with the cognitive inference and learning system. 9. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises a summarization portion, the summarization portion processing information, organizing the information, ranking the information and generating s corresponding summary of the information. 10. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises a temporal/spatial reasoning portion, the temporal/spatial reasoning portion reasoning based upon qualitative abstractions of temporal and spatial aspects of common sense knowledge. 11. The cognitive information processing system environment of claim 1, wherein: the cognitive inference and learning system comprises an entity resolution portion, the entity resolution portion identifying elements in a set of data that refer to the same entity across different data sources. 12. The cognitive information processing system environment of claim 1, wherein: the destination comprises at least one of a mobile application, an alert, a business intelligence type application, a statistical tool, a third party custom application, a marketplace and an application program interface. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A cognitive information processing system environment which includes a plurality of data sources; a cognitive inference and learning system coupled to receive a data from the plurality of data sources, the cognitive inference and learning system processing the data from the plurality of data sources to provide cognitively processed insights, the cognitive inference and learning system further comprising performing a learning operation to iteratively improve the cognitively processed insights over time; and, a destination, the destination receiving the cognitively processed insights. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A cognitive information processing system environment which includes a plurality of data sources; a cognitive inference and learning system coupled to receive a data from the plurality of data sources, the cognitive inference and learning system processing the data from the plurality of data sources to provide cognitively processed insights, the cognitive inference and learning system further comprising performing a learning operation to iteratively improve the cognitively processed insights over time; and, a destination, the destination receiving the cognitively processed insights. |
|
Disclosed herein is a method of managing decision logic. The method includes receiving data, storing the data, and receiving a set of rules. A decision is generated based at least in part on the data and on the set of rules, and is a part of the decision logic. The decision logic is managed in a first mode or a second mode. When in the first mode or the second mode, the set of rules is managed in the context of the data by a first user. The managing includes reviewing and editing the set of rules for the decision logic in the context of the data. The editing is done by at least one of (i) modifying a rule in the set of rules, (ii) creating another rule and adding it to the set of rules and (iii) making an exception to a rule in the set of rules. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A non-transitory computer readable medium storing a program that when executed by a processer performs the method of managing decision logic, the method comprising: receiving data; storing the data in a computer memory; receiving a set of rules for the decision logic; generating a decision based at least in part on the data and on the set of rules, the decision being a part of the decision logic; managing the decision logic in a first mode or a second mode; and when in the first mode or the second mode, managing the set of rules for the decision logic in the context of the data by a first user; wherein the managing includes reviewing the set of rules for the decision logic in the context of the data; and wherein the managing includes editing the set of rules for the decision logic in the context of the data, the editing done by at least one of (i) modifying a rule in the set of rules, (ii) creating another rule and adding it to the set of rules and (iii) making an exception to a rule in the set of rules. 2. The method of claim 1, wherein the first mode is a text based representation including a form or report. 3. The method of claim 1, wherein the second mode is a graphical representation including a table, tree or graph. 4. The method of claim 1, wherein the managing is performed in the first mode or the second mode based on a preference of the first user. 5. The method of claim 1, wherein the first user switches between the first mode and the second mode for managing the decision logic. 6. The method of claim 1, wherein the managing is performed by the first user and a second user at the same time when the first user and the second user are in the same mode. 7. The method of claim 1, wherein the managing is performed by the first user and a second user at the same time when the first user and the second user are in different modes. 8. The method of claim 1, wherein the managing and execution by the first user occurs simultaneously during the method of managing decision logic. 9. The method of claim 1, the method further comprising: when in the first mode, displaying a graphical indicator to the first user, the graphical indicator being associated with the data, the graphical indicator representing information associated with the set of rules of the decision logic. 10. The method of claim 1, the method further comprising: when in the second mode, displaying a graphical indicator to the first user, the graphical indicator being associated with the data, the graphical indicator representing information associated with the set of rules of the decision logic. 11. The method of claim 1, the method further comprising: when in the first mode, receiving a selection of selected contextual data from the user; calculating an impact of the selected contextual data on the decision logic; and displaying a graphical indicator, the graphical indicator showing the impact of the selected contextual data on the decision logic; wherein the impact includes at least one of (i) whether the selected contextual data is used in a rule, the rule being included in the set of rules, (ii) a frequency of usage of the selected contextual data in the decision logic, (iii) a role of the selected contextual data in determining the decision, and (iv) usage information indicating an importance and correlation of the selected contextual data with respect to the decision. 12. The method of claim 1, the method further comprising: when in the second mode, receiving a selection of selected contextual data from the user; calculating an impact of the selected contextual data on the decision logic; and displaying a graphical indicator, the graphical indicator showing the impact of the selected contextual data on the decision logic; wherein the impact includes at least one of (i) whether the selected contextual data is used in a rule, the rule being included in the set of rules, (ii) a frequency of usage of the selected contextual data in the decision logic, (iii) a role of the selected contextual data in determining the decision, and (iv) usage information indicating an importance and correlation of the selected contextual data with respect to the decision. 13. A non-transitory computer readable medium storing a program that when executed by a processer performs the method of investigating decision logic, the method comprising: receiving data; storing the data in a computer memory; receiving a set of rules for the decision logic; generating a decision based at least in part on the data and on the set of rules, the decision being a part of the decision logic; when in a first investigative mode, displaying a first graphical indicator to a user, the first graphical indicator being associated with the data, the first graphical indicator representing information associated with the set of rules of the decision logic; and when in a second investigative mode, displaying a second graphical indicator to the user, the second graphical indicator being associated with the set of rules of the decision logic, the second graphical indicator representing information associated with the data; wherein the user switches between the first investigative mode and the second investigative mode based on a preference of the user; and wherein the second graphical indicator is a graphical rules editor, the second graphical rules editor including at least one of (i) a first option to create a new rule based on the selected contextual data, (ii) a second option to create a new exception rule to an existing rule in the set of rules based on the selected contextual data, and (iii) a third option to modify the existing rule in the set of rules based on the selected contextual data. 14. The method of claim 13, the method further comprising: receiving a selection of selected contextual data from the user; and calculating an impact of the selected contextual data on the decision logic; wherein the first graphical indicator shows the impact of the selected contextual data on the decision logic. 15. The method of claim 14, wherein the impact includes at least one of (i) whether the selected contextual data is used in a rule, the rule being included in the set of rules, (ii) a frequency of usage of the selected contextual data in the decision logic, (iii) a role of the selected contextual data in determining the decision, and (iv) usage information indicating an importance and correlation of the selected contextual data with respect to the decision. 16. The method of claim 13, the method further comprising: receiving a selection of a selected rule from the set of rules from the user; and calculating an impact of the selected rule from the set of rules on the decision logic; wherein the second graphical indicator shows the impact of the selected rule from the set of rules on the decision logic. 17. The method of claim 16, wherein the impact includes at least one of (i) whether the selected rule, the rule being included in the set of rules, is used in the decision logic, (ii) a frequency of usage of the selected rule, the rule being included in the set of rules, in the decision logic, (iii) a role of the selected rule, the rule being included in the set of rules, in determining the decision, and (iv) usage information indicating an importance and correlation of the selected rule, the rule being included in the set of rules, with respect to the decision. 18. The method of claim 16, wherein: the selection includes two rules from the set of rules; and the second graphical indicator is color-coded to show the impact of the rule from the set of rules on the decision logic. 19. The method of claim 13, wherein when in the first investigative mode, the first graphical indicator is a graphical rules editor, the graphical rules editor including at least one of (i) a first option to create a new rule based on the selected contextual data, (ii) a second option to create a new exception rule to an existing rule in the set of rules based on the selected contextual data, and (iii) a third option to modify the existing rule in the set of rules based on the selected contextual data. 20. The method of claim 13, wherein the investigation and execution occurs simultaneously during the method of investigating decision logic. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: Disclosed herein is a method of managing decision logic. The method includes receiving data, storing the data, and receiving a set of rules. A decision is generated based at least in part on the data and on the set of rules, and is a part of the decision logic. The decision logic is managed in a first mode or a second mode. When in the first mode or the second mode, the set of rules is managed in the context of the data by a first user. The managing includes reviewing and editing the set of rules for the decision logic in the context of the data. The editing is done by at least one of (i) modifying a rule in the set of rules, (ii) creating another rule and adding it to the set of rules and (iii) making an exception to a rule in the set of rules. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Disclosed herein is a method of managing decision logic. The method includes receiving data, storing the data, and receiving a set of rules. A decision is generated based at least in part on the data and on the set of rules, and is a part of the decision logic. The decision logic is managed in a first mode or a second mode. When in the first mode or the second mode, the set of rules is managed in the context of the data by a first user. The managing includes reviewing and editing the set of rules for the decision logic in the context of the data. The editing is done by at least one of (i) modifying a rule in the set of rules, (ii) creating another rule and adding it to the set of rules and (iii) making an exception to a rule in the set of rules. |
|
A method includes receiving a set of parameters for a given project and generating, using information from a knowledge database, a plurality of combinations of group members based at least in part on the set of parameters. The method also includes evaluating a set of metrics for each of the combinations of group members, the set of metrics comprising at least one novelty metric and at least one collective intelligence metric. The method further includes generating one or more strategy matrices for each of at least a subset of the combinations of group members using information from the knowledge database, evaluating the combinations of group members in the subset using the strategy matrices to determine respective predicted success values, and selecting a given one of the combinations of group members for the given project based at least in part on the sets of metrics and predicted success values. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method comprising: receiving a set of parameters for a given project; generating, using information from a knowledge database, a plurality of combinations of group members based at least in part on the set of parameters; generating one or more strategy matrices for each of at least a subset of the combinations of group members using information from the knowledge database; evaluating the combinations of group members in the subset using the strategy matrices to determine respective predicted success values; and selecting a given one of the combinations of group members for the given project based at least in part on the predicted success values; wherein the method is performed by at least one processing device comprising a processor coupled to a memory. 2. The method of claim 1, wherein the set of parameters for the given project comprises one or more project goals and one or more group member constraints. 3. The method of claim 2, wherein the set of parameters for the given project further comprises at least one budget constraint. 4. The method of claim 2, wherein the one or more group member constraints specify project roles and a number of group members for each of the project roles. 5. The method of claim 2, wherein the one or more group member constraints specify, for each group member for the given project, at least one of: a minimum number of other group members which that group member has worked with on one or more historical projects to be included in each combination of group members; and a maximum number of other group members which that group member has worked with on one or more historical projects to be included in each combination of group members. 6. The method of claim 1, wherein the knowledge database comprises information associated with a plurality of possible group members comprising one or more of skill levels, previous project roles, estimated costs, behavior metrics and social sensitivity metrics for predicting collective intelligence. 7. The method of claim 1, wherein the knowledge database comprises information associated with a plurality of historical projects comprising one or more of lists of group members, quantified outcomes, project roles and metrics measuring member interactions. 8. The method of claim 1, further comprising evaluating a set of metrics for each of the combinations of group members, the set of metrics being utilized in selecting the given combination of group members for the given project. 9. The method of claim 8, wherein the set of metrics comprises at least one novelty metric comprising a Bayesian surprise metric calculated according to: ∫ ℳ p ( M | A ) log p ( M | A ) p ( M ) M where is a set of groups known to an observer, M∈ and A is a new group being observed and p denotes a probability distribution function. 10. The method of claim 8, wherein the set of metrics comprises at least one collective intelligence metric, the at least one collective intelligence metric comprising psychometrics for objectively measuring skills, knowledge, abilities, attitudes and personality traits. 11. The method of claim 1, wherein the at least one strategy matrix comprises a role-behavior matrix mapping a set of behavior patterns to roles of group members. 12. The method of claim 1, wherein the at least one strategy matrix comprises a role-experience matrix mapping a set of skill levels to roles of group members. 13. The method of claim 1, wherein the at least one strategy matrix comprises a communication matrix mapping sequences of actions to roles of group members. 14. The method of claim 1, wherein evaluating the predicted success value of a combination of group members for the given project comprises using a machine learning algorithm and at least one of the strategy matrices, the machine learning algorithm comprising one or more of a clustering algorithm, a nearest neighbor algorithm, a regression algorithm and a support vector machine (SVM) algorithm. 15. The method of claim 1, further comprising: generating a plurality of communication matrices for the given combination of group members, each communication matrix representing a sequence of actions for the given project corresponding to one or more strategies for the given project; and predicting the success of the one or more strategies for the given project based on a comparison of one or more of the generated communication matrices and one or more historical communication matrices for historical projects. 16. The method of claim 15, wherein predicting the success of the one or more strategies for the given project comprises utilizing a machine learning algorithm and the generated communication matrices. 17. The method of claim 1, wherein the given project comprises a sports game, the group members comprises players on a sports team, the predicted success values comprises game scores and the set of parameters comprises a set of project roles identifying positions for the sports team and a salary cap for the sports team. 18. The method of claim 1, wherein the given project comprise an enterprise project, the group members comprises employees or partners of the enterprise, the predicted success values comprise key performance indicators (KPIs) and the set of parameters comprises a set of project roles identifying job descriptions for the project and a project budget. 19. An article of manufacture comprising a computer readable storage medium for storing computer readable program code which, when executed, causes a computer: to receive a set of parameters for a given project; to generate, using information from a knowledge database, a plurality of combinations of group members based at least in part on the set of parameters; to generate one or more strategy matrices for each of at least a subset of the combinations of group members using information from the knowledge database; to evaluate the combinations of group members in the subset using the strategy matrices to determine respective predicted success values; and to select a given one of the combinations of group members for the given project based at least in part on the predicted success values. 20. An apparatus comprising: at least one processing device comprising a processor coupled to a memory; the at least one processing device being configured: to receive a set of parameters for a given project; to generate, using information from a knowledge database, a plurality of combinations of group members based at least in part on the set of parameters; to generate one or more strategy matrices for each of at least a subset of the combinations of group members using information from the knowledge database; to evaluate the combinations of group members in the subset using the strategy matrices to determine respective predicted success values; and to select a given one of the combinations of group members for the given project based at least in part on the predicted success values. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method includes receiving a set of parameters for a given project and generating, using information from a knowledge database, a plurality of combinations of group members based at least in part on the set of parameters. The method also includes evaluating a set of metrics for each of the combinations of group members, the set of metrics comprising at least one novelty metric and at least one collective intelligence metric. The method further includes generating one or more strategy matrices for each of at least a subset of the combinations of group members using information from the knowledge database, evaluating the combinations of group members in the subset using the strategy matrices to determine respective predicted success values, and selecting a given one of the combinations of group members for the given project based at least in part on the sets of metrics and predicted success values. |
|
G06N3126 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method includes receiving a set of parameters for a given project and generating, using information from a knowledge database, a plurality of combinations of group members based at least in part on the set of parameters. The method also includes evaluating a set of metrics for each of the combinations of group members, the set of metrics comprising at least one novelty metric and at least one collective intelligence metric. The method further includes generating one or more strategy matrices for each of at least a subset of the combinations of group members using information from the knowledge database, evaluating the combinations of group members in the subset using the strategy matrices to determine respective predicted success values, and selecting a given one of the combinations of group members for the given project based at least in part on the sets of metrics and predicted success values. |
|
A method, system and a computer program product are provided for verifying ground truth data by iteratively assigning machine-annotated training set examples to clusters which are prioritized based on verification scores to identify and display one or more prioritized review candidate training set examples grouped in a prioritized cluster in order to solicit verification or correction feedback from a human subject matter expert for inclusion in an accepted training set. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method of verifying ground truth data, the method comprising: receiving, by an information handling system, comprising a processor and a memory, ground truth data comprising a human-curated training set; performing, by the information handling system, annotation operations on the training set using an annotator to generate a machine-annotated training set; assigning, by the information handling system, examples from the machine-annotated training set to one or more clusters according to a feature vector similarity measure; analyzing, by the information handling system, the one or more clusters to prioritize clusters based on verification scores computed for each cluster; and displaying, by the information handling system, machine-annotated training set examples associated with a prioritized cluster as prioritized review candidates to solicit verification or correction feedback from a human subject matter expert (SME) for inclusion in an accepted training set. 2. The method of claim 1, where the annotator comprises a dictionary annotator, rule-based annotator, or a machine learning annotator. 3. The method of claim 1, where assigning examples from the machine-annotated training set to one or more clusters comprises: generating a vector representation for each of example from the machine-annotated training set; and applying one or more feature selection algorithms to the vector representations of the machine-annotated training set examples to identify the one or more clusters. 4. The method of claim 1, where analyzing the one or more clusters comprises computing a verification score for each cluster as a confidence metric which quantifies how likely that annotations in the cluster are true positives based on a training model for the feature set of a given annotation cluster. 5. The method of claim 1, where analyzing the one or more clusters comprises computing a verification score for each cluster as an Inter Annotator Agreement (IAA) score measuring how consistent annotations of the human SME are with annotations from a group of human SMEs for a given annotation cluster. 6. The method of claim 1, where analyzing the one or more clusters comprises computing a verification score for each cluster as a cluster size score measuring a given annotation cluster. 7. The method of claim 1, where analyzing the one or more clusters comprises computing a verification score for each cluster as a cross-validation score measuring how similar the machine-annotated training set examples are to a feature set for a given annotation cluster. 8. The method of claim 1, further comprising verifying or correcting all prioritized review candidates in a cluster as a single group based on verification or correction feedback from the human subject matter expert. 9. A computer program product comprising a computer readable storage medium having a computer readable program stored therein, wherein the computer readable program, when executed on an information handling system, causes the system to verify ground truth data by: receiving ground truth data comprising a human-curated training set; performing annotation operations on the training set using an annotator to generate a machine-annotated training set; assigning examples from the machine-annotated training set to one or more clusters according to a feature vector similarity measure; analyzing the one or more clusters to prioritize clusters based on verification scores computed for each cluster; and displaying machine-annotated training set examples associated with a prioritized cluster as prioritized review candidates to solicit verification or correction feedback from a human subject matter expert (SME) for inclusion in an accepted training set. 10. The computer program product of claim 9, wherein the computer readable program, when executed on the system, causes the system to perform annotation operations using a dictionary annotator, rule-based annotator, or a machine learning annotator. 11. The computer program product of claim 9, wherein the computer readable program, when executed on the system, causes the system to assign examples from the machine-annotated training set to one or more clusters by: generating a vector representation for each of example from the machine-annotated training set; and applying one or more feature selection algorithms to the vector representations of the machine-annotated training set examples to identify the one or more clusters. 12. The computer program product of claim 9, wherein the computer readable program, when executed on the system, causes the system to analyze the one or more clusters by computing a verification score for each cluster as a confidence metric which quantifies how likely that annotations in the cluster are true positives based on a training model for the feature set of a given annotation cluster. 13. The computer program product of claim 9, wherein the computer readable program, when executed on the system, causes the system to analyze the one or more clusters by computing a verification score for each cluster as an Inter Annotator Agreement (IAA) score measuring how consistent annotations of the human SME are with annotations from a group of human SMEs for a given annotation cluster. 14. The computer program product of claim 9, wherein the computer readable program, when executed on the system, causes the system to analyze the one or more clusters by computing a verification score for each cluster as a cluster size score measuring a given annotation cluster. 15. The computer program product of claim 9, wherein the computer readable program, when executed on the system, causes the system to analyze the one or more clusters by computing a verification score for each cluster as a cross-validation score measuring how similar the machine-annotated training set examples are to a feature set for a given annotation cluster. 16. The computer program product of claim 9, further comprising computer readable program, when executed on the system, causes the system to verify or correct all prioritized review candidates in a cluster as a single group based on verification or correction feedback from the human subject matter expert. 17. An information handling system comprising: one or more processors; a memory coupled to at least one of the processors; and a set of instructions stored in the memory and executed by at least one of the processors to verify ground truth data, wherein the set of instructions are executable to perform actions of: receiving, by the system, ground truth data comprising a human-curated training set; perforating, by the system, annotation operations on the training set using an annotator to generate a machine-annotated training set; assigning, by the system, examples from the machine-annotated training set to one or more clusters according to a feature vector similarity measure; analyzing, by the system, the one or more clusters to prioritize clusters based on verification scores computed for each cluster; and displaying, by the system, machine-annotated training set examples associated with a prioritized cluster as prioritized review candidates to solicit verification or correction feedback from a human subject matter expert (SME) for inclusion in an accepted training set. 18. The information handling system of claim 17, where analyzing the one or more clusters comprises computing a verification score for each cluster as a confidence metric which quantifies how likely that annotations in the cluster are true positives based on a training model for the feature set of a given annotation cluster. 19. The information handling system of claim 17, where analyzing the one or more clusters comprises computing a verification score for each cluster as an Inter Annotator Agreement (IAA) score measuring how consistent annotations of the human SME are with annotations from a group of human SMEs for a given annotation cluster. 20. The information handling system of claim 17, where analyzing the one or more clusters comprises computing a verification score for each cluster as a cluster size score measuring a given annotation cluster. 21. The information handling system of claim 17, where analyzing the one or more clusters comprises computing a verification score for each cluster as a cross-validation score measuring how similar the machine-annotated training set examples are to a feature set for a given annotation cluster. 22. The information handling system of claim 17, further comprising verifying or correcting all prioritized review candidates in a cluster as a single group based on verification or correction feedback from the human subject matter expert. 23. The information handling system of claim 17, further comprising verifying or correcting prioritized review candidates in a cluster one at a time based on verification or correction feedback from the human subject matter expert. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method, system and a computer program product are provided for verifying ground truth data by iteratively assigning machine-annotated training set examples to clusters which are prioritized based on verification scores to identify and display one or more prioritized review candidate training set examples grouped in a prioritized cluster in order to solicit verification or correction feedback from a human subject matter expert for inclusion in an accepted training set. |
|
G06N5022 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method, system and a computer program product are provided for verifying ground truth data by iteratively assigning machine-annotated training set examples to clusters which are prioritized based on verification scores to identify and display one or more prioritized review candidate training set examples grouped in a prioritized cluster in order to solicit verification or correction feedback from a human subject matter expert for inclusion in an accepted training set. |
|
Efficient learning of a neural network can be performed. A plurality of DNNs are hierarchically configured, and data of a hidden layer of a DNN of a first hierarchy machine learning/recognizing device is used as input data of a DNN of a second hierarchy machine learning/recognizing device. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An information processing system, comprising: a plurality of DNNs which are hierarchically configured, wherein data of a hidden layer of a DNN of a first hierarchy machine learning/recognizing device is used as input data of a DNN of a second hierarchy machine learning/recognizing device. 2. The information processing system according to claim 1, wherein, after supervised learning is performed in the DNN of the first hierarchy machine learning/recognizing device so that an output layer performs a desired output, supervised learning of the DNN of the second hierarchy machine learning/recognizing device is performed. 3. The information processing system according to claim 1, wherein the first hierarchy machine learning/recognizing device includes a unit that stores a score of a recognition result of a recognition process while performing the recognition process and an update request transmitting unit that transmits an update request signal for a neural network structure and a weight coefficient of the DNN of the first hierarchy machine learning/recognizing device to the second hierarchy machine learning/recognizing device in a case in which the recognition result is larger than a predetermined threshold value 1 or smaller than a predetermined threshold value 2 or in a case in which a variance when a histogram of the recognition result is generated is larger than a predetermined value, upon receiving the update request signal of the first hierarchy machine learning/recognizing device, the second hierarchy machine learning/recognizing device updates the neural network structure and the weight coefficient of the DNN of the first hierarchy machine learning/recognizing device, and transmits update data to the first hierarchy machine learning/recognizing device, and the first hierarchy machine learning/recognizing device constructs a new neural network on the basis of the update data. 4. The information processing system according to claim 1, wherein the first hierarchy machine learning/recognizing device includes a learning module that performs a learning process, a storage unit that stores weight coefficient information of a learning result of the learning process, recognition result rating information, and intermediate layer data information, and a unit that transmits the update request signal to the second hierarchy machine learning/recognizing device in a case in which it is necessary to update the neural network of the first hierarchy machine learning/recognizing device. 5. The information processing system according to claim 1, wherein a connection of the first hierarchy machine learning/recognizing device and the second hierarchy machine learning/recognizing device has only an input from the first hierarchy machine learning/recognizing device to the second hierarchy machine learning/recognizing device. 6. The information processing system according to claim 1, wherein the first hierarchy machine learning/recognizing device includes a storage device that temporarily holds a value of the hidden layer of the DNN and a mechanism that holds data of the storage device in the second hierarchy machine learning/recognizing device as an input data database. 7. The information processing system according to claim 1, wherein there are a plurality of first hierarchy machine learning/recognizing devices, and the plurality of first hierarchy machine learning/recognizing devices are connected directly or via a network using at least one of a wired manner and a wireless manner for transmission of the input data from the plurality of first hierarchy machine learning/recognizing devices to the single second hierarchy machine learning/recognizing device. 8. The information processing system according to claim 1, wherein there are a plurality of second hierarchy machine learning/recognizing devices, and data of the hidden layer data from one of the first hierarchy machine learning/recognizing devices is shared by the plurality of second hierarchy machine learning/recognizing devices. 9. The information processing system according to claim 1, wherein a copy of the DNN of the first hierarchy machine learning/recognizing device is installed in the second hierarchy machine learning/recognizing device, and together with learning or a recognition process in with the first hierarchy machine learning/recognizing device, in the second hierarchy machine learning/recognizing device, learning is performed on the basis of input data from the first hierarchy machine learning/recognizing device, and as a result, configuration information of a neural network and weight coefficient information which are a learning result in the second hierarchy machine learning/recognizing device is transmitted to the first hierarchy machine learning/recognizing device, and the neural network and a weight coefficient of the first hierarchy machine learning/recognizing device are updated. 10. The information processing system according to claim 1, wherein a hardware size of the second hierarchy machine learning/recognizing device is larger than a hardware size of the first hierarchy machine learning/recognizing device. 11. A method for operating an information processing system including a plurality of DNNs, comprising: configuring the plurality of DNNs to have a multi-layer structure including a first hierarchy machine learning/recognizing device and a second hierarchy machine learning/recognizing device; wherein information processing capability of the second hierarchy machine learning/recognizing device higher than information processing capability of the first hierarchy machine learning/recognizing device is used, and data of a hidden layer of a DNN of the first hierarchy machine learning/recognizing device is used as input data of a DNN of the second hierarchy machine learning/recognizing device. 12. The method for operating the information processing system according to claim 11, wherein a configuration of a neural network of the first hierarchy machine learning/recognizing device DNN is controlled on the basis of a processing result of the second hierarchy machine learning/recognizing device. 13. The method for operating the information processing system according to claim 11, wherein one inspection target is observed using a plurality of first hierarchy machine learning/recognizing devices, the data of the hidden layer of the first hierarchy machine learning/recognizing device obtained in a process of the observation is transferred to the second hierarchy machine learning/recognizing device, in the second hierarchy machine learning/recognizing device, learning is performed on the basis of the data of the hidden layer, and a database for calculating a neural network structure and a weight coefficient of the first hierarchy machine learning/recognizing device is constructed, the learning and the construction period of the database in the second hierarchy machine learning/recognizing device are defined as a learning enhancement period of the first hierarchy machine learning/recognizing device, and the second hierarchy machine learning/recognizing device has an operation form of defining an actual operation period in which the neural network and the weight coefficient of the first hierarchy machine learning/recognizing device are set, and an operation of recognition learning is performed in the first hierarchy machine learning/recognizing device and the second hierarchy machine learning/recognizing device after the learning is completed. 14. The method for operating the information processing system according to claim 11, wherein a first learning period for initial neural network construction in the second hierarchy machine learning/recognizing device in order to construct a plurality of first hierarchy machine learning/recognizing devices is set, then, a second learning period in which learning data acquired in the first learning period is loaded to the first hierarchy machine learning/recognizing device, and supervised learning is performed while actually operating the first hierarchy machine learning/recognizing device is set, and further, after the second learning period ends, a third learning period in which machine learning recognition control using the above first hierarchy machine learning/recognizing device is performed, and cooperative learning with the second hierarchy machine learning/recognizing device is performed if necessary is set. 15. A machine learning operator, comprising: a unit that performs an operation on data of a second layer using data of a first layer and performs an operation on data of the first layer using data of the second layer in a multi-layered neural network, wherein weight data of deciding a relation between each piece of data of the first layer and each piece of data of the second layer in both the operations is provided, and the weight data is stored in one storage holding unit as all weight coefficient matrices to be constructed; an operation unit including product-sum operators which are constituent elements of the weight coefficient matrix and correspond to operations of matrix elements in a one-to-one manner, wherein, when the matrix elements constituting the weight coefficient matrix are stored in the storage holding unit, the matrix elements are stored using a row vector of the matrix as a basic unit, the operation of the weight coefficient matrix is performed in basic units in which the storage is performed in the storage holding unit, a first row component of the row vector is held in the storage holding unit so that an arrangement order of constituent elements is the same as a column vector of an original matrix, a second row component of the row vector is held in the storage holding unit after shifting the constituent element of the column vector of the original matrix to the right or the left by one element, a third row component of the row vector is held in the storage holding unit after further shifting the constituent element of the column vector of the original matrix by one element in the same direction as a movement direction in the second row component, and an N-th row component of the last row of the row vector is held in the storage holding unit further shifting the constituent element of the column vector of the original matrix by one element in the same direction as a movement direction in an (N−1)-th row component; and an operator configuration in which, in a case in which the data of the first layer is calculated from the data of the second layer using the weight coefficient matrix, the data of the second layer is arranged similarly to the column vector of the matrix, and each element is input to the product-sum operator, at the same time, a first row of the weight coefficient matrix is input to the product-sum operator, a multiplication operation related to both pieces of data is performed, and an operation result is stored in the accumulator, when second or less rows of the weight coefficient matrix are calculated, the data of the second layer is shifted to the left or the right each time a row operation of the weight matrix is performed, and then a multiplication operation of element data of a corresponding row of the weight coefficient matrix and the arranged data of the second layer is performed, then, data stored in the accumulator of the same operation unit is added, and a similar operation is performed up to an N-th row of the weight coefficient matrix, and in a case in which the data of the second layer is calculated from the data of the first layer using the weight coefficient matrix, the data of the first layer is arranged similarly to the column vector of the matrix, and each element is input to the product-sum operator, at the same time, a first row of the weight coefficient matrix is input to the product-sum operator, a multiplication operation is performed, and a result is stored in the accumulator, when second or less rows of the weight coefficient matrix are calculated, the data of the first layer is shifted to the left or the right each time a row operation of the weight matrix is performed, and then a multiplication operation of element data of a corresponding row of the weight coefficient matrix and the arranged data of the first layer is performed, then, information of the accumulator stored in the operation unit is input to an adding unit of a neighbor operation unit, added to the result of the multiplication operation, and a result is stored in the accumulator, and a similar operation is performed up to the N-th row of the weight matrix. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Efficient learning of a neural network can be performed. A plurality of DNNs are hierarchically configured, and data of a hidden layer of a DNN of a first hierarchy machine learning/recognizing device is used as input data of a DNN of a second hierarchy machine learning/recognizing device. |
|
G06N30454 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Efficient learning of a neural network can be performed. A plurality of DNNs are hierarchically configured, and data of a hidden layer of a DNN of a first hierarchy machine learning/recognizing device is used as input data of a DNN of a second hierarchy machine learning/recognizing device. |
|
Provided is a process, including: obtaining a set of historical geolocations; segmenting the historical geolocations into a plurality of temporal bins; determining pairwise transition probabilities between a set of geographic places based on the historical geolocations; configuring a compute cluster by assigning subsets of the transition probabilities to computing devices in the compute cluster; receiving a geolocation stream indicative of current geolocations of individuals; selecting a computing device in the compute cluster in response to determining that the computing device contain transition probabilities for the received respective geolocation; selecting transition probabilities applicable to the received respective geolocation from among the subset of transition probabilities assigned to the selected computing device; predicting a subsequent geographic place based on the selected transition probabilities. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A system, comprising: one or more processors; and memory storing instructions that when executed by at least some of the processors effectuate operations comprising: training a machine learning model to predict subsequent geolocations of an individual, wherein training the machine learning model comprises: obtaining a set of historical geolocations of more than 500 individuals, the set of historical geolocations indicating geographic places visited by the individuals and the sequence with which the places were visited, the set of historical geolocations indicating a median number of places greater than or equal to three for the 500 individuals over a trailing duration of time extending more than one day in the past; obtaining a set of geographic places including at least 500 geographic places each corresponding to at least one of the historical geolocations; segmenting the historical geolocations into a plurality of temporal bins, each temporal bin having a start time and an end time, wherein segmenting comprises assigning the historical geolocations to a given temporal bin in response to determining that a corresponding historical geolocation is timestamped with a time after a respective start time and before a respective end time of the given temporal bin; and for each of the segments, determining pairwise transition probabilities between the set of geographic places based on the historical geolocations in the respective temporal bin to form a transition matrix, wherein: a first dimension of the transition matrix corresponds to a first previous geographic place, a second dimension of the transition matrix corresponds to a second previous geographic place preceding the first previous geographic place, a third dimension of the transition matrix corresponds to a subsequent geographic place, and values of the transition matrix correspond to respective conditional probabilities of moving from the first previous geographic place to the subsequent geographic place given that the second previous geographic place precedes the first previous geographic place in a sequence of places visited by an individual; configuring a real-time, next-place-prediction stream-processor compute cluster by assigning a first subset of the transition probabilities corresponding to a first computing device in the compute cluster and assigning a second subset of the transition probabilities to a second computing device in the compute cluster; after configuring the stream-processor compute cluster, receiving, with the stream-processor compute cluster, a geolocation stream indicative of current geolocations of individuals, wherein the geolocation stream comprises over 1,000 geolocations per hour; for each geolocation in the stream, within thirty minutes of receiving the respective geolocation, predicting a subsequent geolocation and acting on the prediction, wherein predicting a subsequent geolocation and acting on the prediction comprises: selecting a computing device in the compute cluster in response to determining that the computing device contain transition probabilities for the received respective geolocation; with the selected computing device, selecting transition probabilities applicable to the received respective geolocation from among the subset of transition probabilities assigned to the selected computing device; predicting a subsequent geographic place based on the selected transition probabilities; selecting content based on the predicted subsequent geographic place; and sending the selected content to a mobile computing device that sent the received respective geolocation. 2. The system of claim 1, wherein: a fourth dimension of the transition matrix corresponds to a third previous geographic place preceding the second previous geographic place, and wherein values of the transition matrix correspond to respective conditional probabilities of moving from the first previous geographic place to the subsequent geographic place given that the second previous geographic place precedes the first previous geographic place and given that the third previous geographic place precedes the second previous geographic place in a sequence of places visited by an individual. 3. The system of claim 1, comprising: pruning the transition matrix to remove values that indicate less than a threshold probability and forming a directed, weighted transition graph from the pruned transition matrix. 4. The system of claim 1, wherein training the machine learning model comprises: storing the set of historical geolocations in a distributed file system of another compute cluster and executing a MapReduce routine with the other compute cluster to concurrently determine at least some of the transition probabilities. 5. The system of claim 1, wherein predicting a subsequent geolocation comprises: ingesting the geolocation stream into the compute cluster; and predicting the subsequent geographic place with a data pipeline having parallel branches in a topology specified by data transformations and streams between the data transformations. 6. The system of claim 1, wherein selecting content and sending content are performed predictively, without the content being explicitly requested by a user. 7. The system of claim 1, wherein the transition matrix comprises a fourth dimension corresponding to an attribute of user profiles of individuals corresponding to the historical geolocations, and wherein different transition probabilities are determined for a given transition corresponding to different values of the attribute. 8. The system of claim 1, wherein segmenting the historical geolocations into a plurality of temporal bins comprises segmenting the temporal bins into geospatial bins. 9. The system of claim 1, comprising: pruning values of the matrix that satisfy a threshold; and compressing the pruned matrix into three vectors. 10. The system of claim 1, comprising: compressing the transition matrix into a set of patterns by detecting sequences of visits to geographic places having higher than a threshold probability and consolidating the detected sequences based on attributes of places in the detected sequences. 11. The system of claim 1, wherein selecting a computing device in the compute cluster comprises: parsing a field from a time-stamped tuple containing the received respective geolocation in the stream; and selecting the computing device based on a hash value generated based on the parsed field. 12. The system of claim 1, wherein the matrix encodes a complete trigraph of the set of geographic places. 13. The system of claim 1, wherein selecting a computing device in the computing cluster comprises: instructing the selecting computing device to process at least part of the stream with a remote procedure call implementing a Thrift protocol or a protocol buffer. 14. The system of claim 1, wherein predicting a subsequent geographic place comprises: predicting a subsequent geographic place based on a state of a neuron in a directed cycle of an artificial neural network. 15. The system of claim 1, wherein predicting a subsequent geographic place comprises: determining which of the selected transition probabilities has a highest probability among the selected probabilities; and determining that the subsequent geographic place is a geographic place corresponding to a position in the third dimension of the transition matrix. 16. The system of claim 1, wherein predicting a subsequent geographic place comprises: retrieving from memory a previous place visited by an individual corresponding to the received respective geolocation; retrieving from memory of the selected computing device transition probabilities corresponding to a sequence of visits including both the previous place and the received respective geolocation. 17. The system of claim 1, wherein predicting a subsequent geographic place comprises: selecting a subsequent place of interest according to both a transition probability and a response rate associated with content pertaining to the subsequent place of interest. 18. The system of claim 1, wherein predicting a subsequent geographic place comprises: predicting the subsequent geographic place with means for predicting the subsequent geographic place. 19. The system of claim 1, wherein selecting content corresponding to the predicted subsequent geographic place comprises: selecting content corresponding to the predicted subsequent geographic place with means for selecting content corresponding to the predicted subsequent geographic place. 20. The system of claim 1, wherein selecting content comprises: selecting content based on the predicted subsequent geographic place and other attributes including either user affinity or content popularity. 21. A method, comprising: training a machine learning model to predict subsequent geolocations of an individual, wherein training the machine learning model comprises: obtaining a set of historical geolocations of more than 500 individuals, the set of historical geolocations indicating geographic places visited by the individuals and the sequence with which the places were visited, the set of historical geolocations indicating a median number of places greater than or equal to three for the 500 individuals over a trailing duration of time extending more than one day in the past; obtaining a set of geographic places including at least 500 geographic places each corresponding to at least one of the historical geolocations; segmenting the historical geolocations into a plurality of temporal bins, each temporal bin having a start time and an end time, wherein segmenting comprises assigning the historical geolocations to a given temporal bin in response to determining that a corresponding historical geolocation is timestamped with a time after a respective start time and before a respective end time of the given temporal bin; and for each of the segments, determining pairwise transition probabilities between the set of geographic places based on the historical geolocations in the respective temporal bin to form a transition matrix, wherein: a first dimension of the transition matrix corresponds to a first previous geographic place, a second dimension of the transition matrix corresponds to a second previous geographic place preceding the first previous geographic place, a third dimension of the transition matrix corresponds to a subsequent geographic place, and values of the transition matrix correspond to respective conditional probabilities of moving from the first previous geographic place to the subsequent geographic place given that the second previous geographic place precedes the first previous geographic place in a sequence of places visited by an individual; configuring a real-time, next-place-prediction stream-processor compute cluster by assigning a first subset of the transition probabilities corresponding to a first computing device in the compute cluster and assigning a second subset of the transition probabilities to a second computing device in the compute cluster; after configuring the stream-processor compute cluster, receiving, with the stream-processor compute cluster, a geolocation stream indicative of current geolocations of individuals, wherein the geolocation stream comprises over 1,000 geolocations per hour; for each geolocation in the stream, within thirty minutes of receiving the respective geolocation, predicting a subsequent geolocation and acting on the prediction, wherein predicting a subsequent geolocation and acting on the prediction comprises: selecting a computing device in the compute cluster in response to determining that the computing device contain transition probabilities for the received respective geolocation; with the selected computing device, selecting transition probabilities applicable to the received respective geolocation from among the subset of transition probabilities assigned to the selected computing device; predicting a subsequent geographic place based on the selected transition probabilities; selecting content based on the predicted subsequent geographic place; and sending the selected content to a mobile computing device that sent the received respective geolocation. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Provided is a process, including: obtaining a set of historical geolocations; segmenting the historical geolocations into a plurality of temporal bins; determining pairwise transition probabilities between a set of geographic places based on the historical geolocations; configuring a compute cluster by assigning subsets of the transition probabilities to computing devices in the compute cluster; receiving a geolocation stream indicative of current geolocations of individuals; selecting a computing device in the compute cluster in response to determining that the computing device contain transition probabilities for the received respective geolocation; selecting transition probabilities applicable to the received respective geolocation from among the subset of transition probabilities assigned to the selected computing device; predicting a subsequent geographic place based on the selected transition probabilities. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Provided is a process, including: obtaining a set of historical geolocations; segmenting the historical geolocations into a plurality of temporal bins; determining pairwise transition probabilities between a set of geographic places based on the historical geolocations; configuring a compute cluster by assigning subsets of the transition probabilities to computing devices in the compute cluster; receiving a geolocation stream indicative of current geolocations of individuals; selecting a computing device in the compute cluster in response to determining that the computing device contain transition probabilities for the received respective geolocation; selecting transition probabilities applicable to the received respective geolocation from among the subset of transition probabilities assigned to the selected computing device; predicting a subsequent geographic place based on the selected transition probabilities. |
|
Methods and systems for determining past activities of a user and triggering an action accordingly are disclosed. In one aspect, the method is performed on a device and involves determining a set of historical information related to past activities of the user, according to at least one sources of historical data available through the device, determining a timeline of past activities of the user, according to the set of historical information, and triggering an action on the device, according to the timeline of past activities of the user. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for triggering, on a device, an action according to past activities of a user, comprising: during a first step, determining a set of historical information related to past activities of the user, according to at least one sources of historical data available through the device; during a second step, determining a timeline of past activities of the user, according to the set of historical information; and during a third step, triggering an action on the device, according to the timeline of past activities of the user; wherein during the second step, the timeline of the past activities of the user is determined by: determining a first list of important places where the user went previously, according to the set of historical information; determining a second list of likely complete visits for which at least an associated period and at least an associated location have been identified, according to the set of historical information; determining a third list of candidate visits for which at least a hint of an associated period and at least a hint of an associated location have been identified, according to the set of historical information; for each candidate visit of the third list: obtaining a candidate place associated to said candidate visit; searching for an important place of the first list matching the candidate place; and if an important place of the first list matches the candidate place, replacing, in said candidate visit, the candidate place by said important place, and moving said candidate visit from the third list to the second list of likely complete visits; and determining the timeline of the past activities of a user, according to the second list of likely complete visits. 2. The method of claim 1, wherein the at least one source of historical data comprises a source of time based events, stored by or accessible through the device. 3. The method of claim 2, wherein the historical information related to past activities of the user determined according to the source of time based events is filtered so as to keep only at least one of: historical data extracted from a personal calendar of the user; and historical data comprising information related to a location. 4. The method of claim 2, wherein the historical information related to past activities of the user determined according to the source of time based events is filtered so as to keep only historical data comprising information related to a location and meeting a criteria based on the likeliness that the user participated to the corresponding event. 5. The method of claim 1, wherein the at least one source of historical data is a source of transaction records, stored by or accessible through the device. 6. The method of claim 5, wherein the historical information related to past activities of the user determined according to the source transaction records is filtered so as to keep only historical data comprising information related to a location. 7. The method of claim 1, wherein the at least one source of historical data is a source of media, stored by or accessible through the device. 8. The method of claim 7, wherein the historical information related to past activities of the user determined according to the source of media is filtered so as to keep only historical data related to media produced by the device or another device associated to the user. 9. The method of claim 1, wherein the at least one source of historical data is a source of informal communications, stored by or accessible through the device. 10. The method of claim 9, wherein the historical information related to past activities of the user determined according to the source of informal communication is filtered so as to keep only historical data comprising information related to a location and a date. 11. A non-transitory computer readable storage medium comprising a set of instructions which, when executed by a processor, cause the processor to implement the method of claim 1. 12. A device, comprising: means for determining a first set of historical information related to past activities of a user, according to at least one source of historical data through the device; means for determining a timeline of past activities of the user, according to the first set of historical information; and means for triggering an action on the device, according to the timeline of past activities of the user; wherein the means for determining a timeline of past activities of the user, according to the first set of historical information, are configured to: determine a first list of important places where the user went previously, according to the set of historical information; determine a second list of likely complete visits for which at least an associated period and at least an associated location have been identified, according to the set of historical information; determine a third list of candidate visits for which at least a hint of an associated period and at least a hint of an associated location have been identified, according to the set of historical information; for each candidate visit of the third list: obtain a candidate place associated to said candidate visit; search for an important place of the first list matching the candidate place; and if an important place of the first list matches the candidate place, replace, in said candidate visit, the candidate place by said important place, and move said the candidate visit from the third list to the second list of likely complete visits; and determine the timeline of the past activities of a user, according to the second list of likely complete visits. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Methods and systems for determining past activities of a user and triggering an action accordingly are disclosed. In one aspect, the method is performed on a device and involves determining a set of historical information related to past activities of the user, according to at least one sources of historical data available through the device, determining a timeline of past activities of the user, according to the set of historical information, and triggering an action on the device, according to the timeline of past activities of the user. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Methods and systems for determining past activities of a user and triggering an action accordingly are disclosed. In one aspect, the method is performed on a device and involves determining a set of historical information related to past activities of the user, according to at least one sources of historical data available through the device, determining a timeline of past activities of the user, according to the set of historical information, and triggering an action on the device, according to the timeline of past activities of the user. |
|
A machine-learning based artificial intelligence device for finding an estimate of patent quality, such as patent lifetime or term is disclosed. Such a device may receive a first set of patent data and generate a list of binary classifiers. A candidate set of binary classifiers may be selected and using a heuristic search, for example an artificial neural network (ANN), a genetic algorithm, a final set of binary classifiers is found by maximizing iteratively a yield according to a cost function, such an area under a curve (AUC) of a receiver operating characteristic (ROC). The device may then receive patent information for a target patent and report an estimate of patent quality according to the final set of binary classifiers. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A machine-learning based artificial intelligence device for finding an estimate of patent quality, the device comprising: a patent data retriever configured to receive a first set of patent data comprising at least one of patent application data and patent data for a plurality of patents, and to generate a list of binary classifiers based on the first set of patent data; a quantitative data scalar configured to assign a standardized scaled score to each binary classifier of the list of binary classifiers; and a binary classifier optimizer configured to generate, using an automated processor, a candidate set of binary classifiers from the list of binary classifiers using a heuristic search and to generate, using the automated processor, a final set of binary classifiers by maximizing iteratively a yield according to a cost function, wherein the device is configured to provide a signal representing the final set of binary classifiers. 2. The device of claim 1, wherein the heuristic search comprises an artificial neural network model. 3. The device of claim 2, wherein the maximizing iteratively comprises changing a number of hidden layers of the artificial neural network. 4. The device of claim 1, wherein the maximizing iteratively comprises using a genetic algorithm. 5. The device of claim 1, wherein the maximizing iteratively comprises using an artificial neural network model and a genetic algorithm. 6. The device of claim 1, wherein the cost function is a receiver operating characteristic and the yield is calculated according an area under a curve. 7. The device of claim 1, wherein the estimate of patent quality represents an estimate of a lifetime of the patent. 8. The device of claim 1, wherein the patent data retriever is configured to receive a second set of patent data comprising at least one of patent application data and patent data for a plurality of patents, and wherein the device is configured to test a validity of the final set of binary classifiers using the second set of patent data. 9. The device of claim 1, further comprising a user information manager configured to receive patent information for a target patent and to report the estimate of patent quality according to the final set of binary classifiers. 10. A system comprising the device of claim 1 and a second device communicatively connected to the device over a network, the second device comprising: a second automated processor: a user interface receiving the patent information for the target patent; an estimate requester requesting from the device the estimate of patent quality for the target patent; and the user interface providing to a user a signal representing the estimate of patent quality. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A machine-learning based artificial intelligence device for finding an estimate of patent quality, such as patent lifetime or term is disclosed. Such a device may receive a first set of patent data and generate a list of binary classifiers. A candidate set of binary classifiers may be selected and using a heuristic search, for example an artificial neural network (ANN), a genetic algorithm, a final set of binary classifiers is found by maximizing iteratively a yield according to a cost function, such an area under a curve (AUC) of a receiver operating characteristic (ROC). The device may then receive patent information for a target patent and report an estimate of patent quality according to the final set of binary classifiers. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A machine-learning based artificial intelligence device for finding an estimate of patent quality, such as patent lifetime or term is disclosed. Such a device may receive a first set of patent data and generate a list of binary classifiers. A candidate set of binary classifiers may be selected and using a heuristic search, for example an artificial neural network (ANN), a genetic algorithm, a final set of binary classifiers is found by maximizing iteratively a yield according to a cost function, such an area under a curve (AUC) of a receiver operating characteristic (ROC). The device may then receive patent information for a target patent and report an estimate of patent quality according to the final set of binary classifiers. |
|
Some embodiments provide systems and methods for enabling a learning implicit gesture control system for use by an occupant of a vehicle. The method includes identifying features received from a plurality of sensors and comparing the features to antecedent knowledge stored in memory. A system output action that corresponds to the features can then be provided in the form of a first vehicle output. The method further includes detecting a second vehicle output from the plurality of sensors and updating the antecedent knowledge to associate the system output action with the second vehicle output. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for enabling a learning implicit gesture control by an occupant of a vehicle, comprising: identifying a first set of features in a plurality of inputs received from a plurality of sensors; comparing the first set of features to an antecedent knowledge stored in a memory device to identify a system output action corresponding to the first set of features and a first vehicle output, the system output action being configured to cause a first vehicle output; detecting a second vehicle output from the plurality of inputs; and updating the antecedent knowledge to associate the system output action with the second vehicle output. 2. The method of claim 1, including only updating the antecedent knowledge when the second vehicle output occurs within a pre-determined time period of the system output action being identified. 3. The method of claim 1, wherein the antecedent knowledge includes a plurality of previously extracted and stored features associated with a given system output. 4. The method of claim 1, wherein the antecedent knowledge includes knowledge stored from deep learning through a generalist network, a specialist network, and a monitor network. 5. The method of claim 1, wherein the plurality of sensors include visual sensors. 6. The method of claim 1, wherein the plurality of sensors include audio sensors. 7. The method of claim 1, wherein the system output action includes modifying a status of at least one of a plurality of systems of the vehicle. 8. The method of claim 7, wherein the plurality of systems of the vehicle include at least one of a reading light, a navigation system, and a telecommunications system. 9. A method for enabling learning implicit gesture control by an occupant of a vehicle, comprising: identifying a first set of features in a plurality of inputs received from a plurality of sensors; inputting the first set of features to a deep knowledge learning system stored in a memory device to identify a system output action corresponding to the first set of features and a first vehicle output, the system output action being configured to cause a first vehicle output; detecting a second vehicle output from the plurality of inputs; and when the second vehicle output occurs within a pre-determined time period of the system output action being generated, performing a self-healing action to update the antecedent knowledge. 10. The method of claim 9, wherein the self-healing action includes updating the antecedent knowledge to associate the system output action with the second vehicle output. 11. The method of claim 10, including, before updating the antecedent knowledge, prompting the user to confirm that the system output action should be associated with the second vehicle output. 12. The method of claim 9, wherein the self-healing action includes distinguishing the first set of features from a second set of features in the antecedent knowledge, the second set of features in the antecedent knowledge being associated with the system output action configured to cause the first vehicle output. 13. An implicit gesture learning system, the system comprising: a processor, the processor in communication with a plurality of sensors coupled to a vehicle; the processor being configured to perform the steps of: identifying a first set of features in a plurality of inputs received from the plurality of sensors; comparing the first set of features to an antecedent knowledge stored in a memory device to identify a system output action corresponding to the first set of features and a first vehicle output, the system output action being configured to cause a first vehicle output; detecting a second vehicle output from the plurality of inputs; and updating the antecedent knowledge to associate the system output action with the second vehicle output. 14. The system of claim 13, wherein the processor is configured to only update the antecedent knowledge when the second vehicle output occurs within a pre-determined time period of the system output action being generated. 15. The system of claim 13, wherein the antecedent knowledge includes a plurality of previously extracted and stored features associated with a given system output. 16. The system of claim 13, wherein the antecedent knowledge includes knowledge stored from deep learning through a generalist network, a specialist network, and a monitor network. 17. The system of claim 13, wherein the plurality of sensors include visual sensors. 18. The system of claim 13, wherein the plurality of sensors include audio sensors. 19. The system of claim 13, wherein the system output action includes modifying a status of at least one of a plurality of systems of the vehicle. 20. The system of claim 19, wherein the plurality of systems of the vehicle include at least one of a reading light, a navigation system, and a telecommunications system. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: Some embodiments provide systems and methods for enabling a learning implicit gesture control system for use by an occupant of a vehicle. The method includes identifying features received from a plurality of sensors and comparing the features to antecedent knowledge stored in memory. A system output action that corresponds to the features can then be provided in the form of a first vehicle output. The method further includes detecting a second vehicle output from the plurality of sensors and updating the antecedent knowledge to associate the system output action with the second vehicle output. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Some embodiments provide systems and methods for enabling a learning implicit gesture control system for use by an occupant of a vehicle. The method includes identifying features received from a plurality of sensors and comparing the features to antecedent knowledge stored in memory. A system output action that corresponds to the features can then be provided in the form of a first vehicle output. The method further includes detecting a second vehicle output from the plurality of sensors and updating the antecedent knowledge to associate the system output action with the second vehicle output. |
|
A machine learning module may generate a probability distribution from training data including labeled modeling data correlated with reflection data. Modeling data may include data from a LIDAR system, camera, and/or a GPS for a target environment/object. Reflection data may be collected from the same environment/object by a radar and/or an ultrasonic system. The probability distribution may assign reflection coefficients for radar and/or ultrasonic systems conditioned on values for modeling data. A mapping module may create a reflection model to overlay a virtual environment assembled from a second set of modeling data by applying the second set to the probability distribution to assign reflection values to surfaces within the virtual environment. Additionally, a test bench may evaluate an algorithm, for processing reflection data to generate control signals to an autonomous vehicle, with simulated reflection data from a virtual sensor engaging reflection values assigned within the virtual environment. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A system, comprising: a set of training data, stored on a physical media set, comprising: three-dimensional modeling data from at least one of a camera system, a LIght-Detection-And-Ranging (LIDAR) system, and a locational-awareness system; and physics modeling data from at least one of a radar system and an ultrasonic system; and a machine learning module, implemented on a processor set, operable to learn a mapping to physical modeling characteristics from three-dimensional modeling information. 2. The system of claim 1, further comprising: a mapping module, on the processor set, operable to create a physical model for the virtual environment by applying the mapping to the three-dimensional modeling information, the physical model providing reflection values for at least one of a virtual radar and a virtual ultrasonic system in the virtual environment; and wherein: the three-dimensional modeling information, stored on the physical media set, carries information with which to generate a three-dimensional model of a virtual environment. 3. The system of claim 2, further comprising: a model generation module, on the processor set, operable to assemble the three-dimensional modeling information into the three-dimensional model of the virtual environment; and wherein the mapping module is further operable to apply the mapping to the three-dimensional modeling information by assigning reflectivity coefficients to surfaces in the three-dimensional model of the virtual environment. 4. The system of claim 2, further comprising a test bench, implemented on the processor set, operable to implement a test algorithm over a path through the virtual environment, the test algorithm processing reflection values received from the virtual environment throughout the path and providing outcomes in the form of at least one of detections, classifications, and determinations. 5. The system of claim 2, further comprising a set of test benches, implemented on the processor set, comprising: at least one Hardware-In-Loop (HIL) system operable to provide mathematical models for mechanical dynamics of a virtual vehicle as it travels a path through the virtual environment and dynamics for the at least one of the virtual radar and the virtual ultrasonic system; and a Software-In-Loop (SIL) system communicably coupled to the at least one HIL system and operable to implement a test algorithm over the path through the virtual environment, the test algorithm processing reflection values received from the virtual environment throughout the path and providing control signals to the virtual vehicle. 6. The system of claim 1, further comprising: a model generation module, on the processor set, operable to assemble the three-dimensional modeling information into a three-dimensional model of a virtual object; and wherein the mapping module is further operable to apply the mapping to the three-dimensional modeling information by assigning reflectivity coefficients to surfaces of the three-dimensional model of the virtual object, creating the virtual object for positioning in a virtual environment with a reflection cross-section for at least one of the virtual radar and the virtual ultrasonic system relative to the virtual object. 7. The system of claim 1, wherein: the set of training data is indexed to labels for a set of target categories; and the machine learning module comprises a deep neural network operable to be trained with supervised learning using the labels indexed to the set of training data. 8. The system of claim 7, wherein the deep neural network further comprises a classifier implemented with a convolution neural network. 9. The system of claim 1, further comprising a heuristic module within the machine learning module operable to identify aspects of the three-dimensional modeling data subject to a set of heuristics informing the machine learning module. 10. A method for physical modeling, further comprising: applying a machine learning algorithm to a set of training data to create a probability distribution for reflection coefficients conditioned on modeling data acquired by at least one of a camera system, a LIght-Detection-And-Ranging (LIDAR) system, and a position system; assigning reflection-coefficient values to a set of modeling data by applying the set of modeling data to the probability distribution. 11. The method of claim 10, further comprising: assembling a set of surface regions from the set of modeling data to model a simulation environment simulating an environment from which the set of modeling data is captured; the step of assigning reflection-coefficient values further comprises: assigning reflection-coefficient values to the set of surface regions; collecting reflection data from transmissions from a virtual sensor placed within the simulation environment, as reflected in accordance with reflection-coefficient values assigned to the set of surface regions in the simulation environment; and testing a perception algorithm on the reflection data. 12. The method of claim 11, wherein testing the perception algorithm further comprises: placing a virtual vehicle within the simulation environment, coupled to the virtual sensor and operable to travel a path through the simulation environment; and sending control signals output by the perception algorithm to the virtual vehicle. 13. The method of claim 10, further comprising: assembling a set of surface regions from the set of modeling data to form a virtual object; the step of assigning reflection-coefficient values further comprises assigning reflection-coefficient values to the set of surface regions; placing the virtual object within a virtual environment; collecting reflection data from transmissions from a virtual sensor placed within the virtual environment, as reflected in accordance with reflection-coefficient values assigned to the set of surface regions in the simulation environment; and testing a perception algorithm on the reflection data. 14. The method of claim 10, wherein: the step of applying a machine learning algorithm to a set of training data further comprises performing supervised learning on a deep neural network with the training data; and after training, the deep neural network implements the probability distribution. 15. The method of claim 10, further comprising tagging portions of at least one of the set of training data and the set of modeling data, which portions correspond to regions in at least one environment from which the at least one of the set of training data and the set of modeling data is collected, and which regions are correlated with ranges of reflection-coefficient values. 16. The method of claim 10, further comprising collecting the set of training data, which further comprises: collecting a training set of modeling data for a set of target areas; collecting a training set of reflection data from the set of target areas correlated with the training set of modeling data; identifying aspects of the set of target areas in the training data with a set of labels for supervised learning with the set of training data. 17. A system for modeling reflections, comprising: a set of virtual-environment data, stored on a physical media set, captured by at least one of a camera system, a Light-Detection-And-Ranging (LIDAR) system, and a locational-awareness system, sufficient to generate a three-dimensional model of a virtual environment; and a mapping module, implemented on a processer set, operable to create a reflection model for the virtual environment by applying the virtual-environment data to a distribution function assigning reflection values to surfaces within the three-dimensional model of the virtual environment, for at least one of a virtual radar system and a virtual ultrasonic system. 18. The system of claim 17, further comprising a machine learning module operable to generate the probability distribution from training data comprising labeled modeling data correlated with reflection data from at least one of a radar system and an ultrasonic system. 19. The system of claim 18, further comprising a heuristic model operable to provide information to the machine learning module about a location relative to an overall field of view of a sensor, used in capturing the training data, from which a portion of the training data originates. 20. The system of claim 17, further comprising a test bench operable to evaluate a reflection-processing algorithm with simulated reflection data obtained from a virtual sensor engaging reflection values assigned to surfaces of the three-dimensional model encountered by the virtual sensor traveling within the virtual environment. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A machine learning module may generate a probability distribution from training data including labeled modeling data correlated with reflection data. Modeling data may include data from a LIDAR system, camera, and/or a GPS for a target environment/object. Reflection data may be collected from the same environment/object by a radar and/or an ultrasonic system. The probability distribution may assign reflection coefficients for radar and/or ultrasonic systems conditioned on values for modeling data. A mapping module may create a reflection model to overlay a virtual environment assembled from a second set of modeling data by applying the second set to the probability distribution to assign reflection values to surfaces within the virtual environment. Additionally, a test bench may evaluate an algorithm, for processing reflection data to generate control signals to an autonomous vehicle, with simulated reflection data from a virtual sensor engaging reflection values assigned within the virtual environment. |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A machine learning module may generate a probability distribution from training data including labeled modeling data correlated with reflection data. Modeling data may include data from a LIDAR system, camera, and/or a GPS for a target environment/object. Reflection data may be collected from the same environment/object by a radar and/or an ultrasonic system. The probability distribution may assign reflection coefficients for radar and/or ultrasonic systems conditioned on values for modeling data. A mapping module may create a reflection model to overlay a virtual environment assembled from a second set of modeling data by applying the second set to the probability distribution to assign reflection values to surfaces within the virtual environment. Additionally, a test bench may evaluate an algorithm, for processing reflection data to generate control signals to an autonomous vehicle, with simulated reflection data from a virtual sensor engaging reflection values assigned within the virtual environment. |
|
Computer-implemented systems and methods are disclosed for automatically generating predictive models using data driven featurization. The systems and methods provide for obtaining data associated with a target event, annotating the data to identify a target event and establishing one or more limits on the data, censoring the data based on the annotations, determining features of the censored data, and analyzing the features to determine a predictive model. In some embodiments, the systems and methods further provide for converting the features into a binary representation and analyzing the binary representation to produce the predictive model. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An electronic device comprising: a data storage configured to store one or more data sets associated with a target event; a data input engine configured to obtain the one or more data sets associated with the target event from one or more data sources; an annotator configured to annotate the one or more data sets, wherein the annotations include an identification of records of the one or more data sets that are associated with the target event; a data censor configured to censor the one or more data sets based on the annotations; and a summarizer configured to determine one or more features of the censored data; an analysis engine configured to: analyze the one or more features, wherein the analysis identifies a subset of the one or more features that are indicative of the target event; and produce a predictive model based on the analysis. 2. The electronic device of claim 1, further comprising a booleanizer configured to convert the features into a binary representation of the features and the analysis engine is further configured to analyze the binary representation. 3. The electronic device of claim 2, wherein conversion of features into the binary representation uses quantile binning. 4. The electronic device of claim 2, wherein the analysis of the binary representation includes chi-squared modeling. 5. The electronic device of claim 1, further comprising a feedback engine configured to obtain feedback associated with the predictive model. 6. The electronic device of claim 5, wherein the feedback engine obtains feedback from domain experts. 7. A method performed by one or more processors and comprising: obtaining one or more data sets associated with a target event from one or more data sources; annotating the one or more data sets, wherein the annotations include identifying the target event within the one or more data sets, identifying records of the one or more data sets that are associated with the target event, and establishing one or more limits on the one or more data sets; censoring the one or more data sets based on the annotations; determining one or more features of the censored data, wherein the one or more features are representative of data in the one or more data sets; analyzing the one or more features, wherein the analysis identifies a subset of the one or more features that are indicative of the target event; and producing a predictive model based on the analysis. 8. The method of claim 7, further comprising converting the one or more features into a binary representation and analyzing the binary representation. 9. The method of claim 8, wherein converting the o or snore features into a binary representation uses quantile binning. 10. The method of claim 8, wherein analyzing the binary representation uses chi-squared modeling. 11. The method of claim 7, further comprising obtaining feedback associated with the predictive model. 12. The method of claim 11, wherein the feedback is obtained from domain experts. 13. A non-transitory computer readable storage medium storing a set of instructions that are executable by a first computing device that includes one or more processors to cause the first computing device to perform a method for evaluating costs associated with a first event, the method comprising: obtaining one or more data sets associated with a target event from one or more data sources; annotating the one or more data sets, wherein the annotations include identifying the target event within the one or more data sets, identifying records of the one or more data sets that are associated with the target event, and establishing one or more limits on the one or more data sets; censoring the one or more data sets based on the annotations; determining one or more features of the censored data, wherein the one or more features are representative of data in the one or more data sets; analyzing the one or more features, wherein the analysis identifies a subset of the one or more features that are indicative of the target event; and producing a predictive model based on the analysis. 14. The non-transitory computer-readable storage medium of claim 13, wherein the set of instructions that are executable by the at least one processor of the first computing device cause the first computing device to further perform: converting the one or more features into a binary representation and analyzing the binary representation. 15. The non-transitory computer-readable storage medium of claim 14, wherein converting the one or more features into a binary representation uses quantile binning. 16. The non-transitory computer-readable storage medium of claim 14, wherein analyzing the binary representation uses chi-squared modeling. 17. The non-transitory computer-readable storage medium of claim 13, wherein the set of instructions that are executable by the at least one processor of the first computing device to cause the first computing device to further perform: obtaining feedback associated with the predictive model. 18. The non-transitory computer-readable storage medium of claim 17, wherein the feedback is obtained from domain experts. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Computer-implemented systems and methods are disclosed for automatically generating predictive models using data driven featurization. The systems and methods provide for obtaining data associated with a target event, annotating the data to identify a target event and establishing one or more limits on the data, censoring the data based on the annotations, determining features of the censored data, and analyzing the features to determine a predictive model. In some embodiments, the systems and methods further provide for converting the features into a binary representation and analyzing the binary representation to produce the predictive model. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Computer-implemented systems and methods are disclosed for automatically generating predictive models using data driven featurization. The systems and methods provide for obtaining data associated with a target event, annotating the data to identify a target event and establishing one or more limits on the data, censoring the data based on the annotations, determining features of the censored data, and analyzing the features to determine a predictive model. In some embodiments, the systems and methods further provide for converting the features into a binary representation and analyzing the binary representation to produce the predictive model. |
|
Optimized learning settings of neural networks are efficiently determined by an apparatus including a processor and one or more computer readable mediums collectively including instructions that, when executed by the processor, cause the processor to train a first neural network with a learning setting; extract tentative weight data from the first neural network with the learning setting; calculate an evaluation value of the first neural network with the learning setting; and generate a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An apparatus comprising: a processor; and one or more computer readable mediums collectively including instructions that, when executed by the processor, cause the processor to: train a first neural network with a learning setting; extract tentative weight data from the first neural network with the learning setting; calculate an evaluation value of the first neural network with the learning setting; and generate a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network. 2. The apparatus of claim 1, wherein the calculating the evaluation value calculates the evaluation value of the first neural network with the learning setting and weight data further trained from the tentative weight data. 3. The apparatus of claim 2, wherein the instructions further cause the processor to: train the second neural network with the new setting; and estimate the evaluation value of the second neural network with the new setting by using the predictive model before completion of training the second neural network with the new setting. 4. The apparatus of claim 3, wherein the instructions further cause the processor to: terminate the training of the second neural network with the new setting in response to the evaluation value of the second neural network with the new setting not satisfying a criterion. 5. The apparatus of claim 4, wherein the instructions further cause the processor to: generate a plurality of new settings; train a plurality of neural networks, each neural network including a respective new setting among the plurality of new settings; terminate the training of at least one neural network among the plurality of neural networks that does not satisfy the criterion; and select one setting based on performances of neural networks of which training is not terminated. 6. The apparatus of claim 5, wherein the instructions further cause the processor to: update the predictive model based on the neural networks of which training is not terminated. 7. The apparatus of claim 2, wherein the training of the first neural network with the learning setting includes a plurality of iterations, and wherein the tentative weight data of the first neural network with the learning setting is updated in each of the plurality of iterations. 8. The apparatus of claim 7, wherein the generation of the predictive model includes generating a function to estimate the evaluation value from the tentative weight data at two or more iterations of the plurality of iterations. 9. The apparatus of claim 8, wherein the two or more iterations are not consecutive. 10. The apparatus of claim 8, wherein the function is operable to estimate the evaluation value from differences between the tentative weight data of a first iteration of the plurality of iterations and the tentative weight data of a second iteration of the plurality of iterations. 11. The apparatus of claim 10, wherein generating the predictive model further normalizes the tentative weight data of a first iteration of the plurality of iterations and the tentative weight data of a second iteration of the plurality of iterations. 12. The apparatus of claim 1, wherein generating the predictive model further normalizes the difference between the tentative weight data of a first iteration of the plurality of iterations and the tentative weight data of a second iteration of the plurality of iterations. 13. The apparatus of claim 12, wherein the tentative weight data is extracted only from the last convolutional layer. 14. The apparatus of claim 1, wherein the first and second neural networks are convolutional neural networks, and at least part of the tentative weight data is extracted from a last convolutional layer. 15. A computer-implemented method comprising: training a first neural network with a learning setting; extracting tentative weight data from the first neural network with the learning setting; calculating an evaluation value of the first neural network with the learning setting; and generating a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network. 16. The computer-implemented method of claim 15, wherein the calculating the evaluation value including calculating the evaluation value of the first neural network with the learning setting and weight data further trained from the tentative weight data. 17. The computer-implemented method of claim 16, further comprising: training the second neural network with the new setting; and estimating the evaluation value of the second neural network with the new setting by using the predictive model before completion of training the second neural network with the new setting. 18. The computer-implemented method of claim 17, further comprising: terminating the training of the second neural network with the new setting in response to the evaluation value of the second neural network with the new setting not satisfying a criterion. 19. The computer-implemented method of claim 18, further comprising: generating a plurality of new settings; training a plurality of neural networks, each neural network including a respective new setting among the plurality of new settings; terminating the training of at least one neural network among the plurality of neural networks that does not satisfy the criterion; and selecting one setting based on performances of neural networks of which training is not terminated. 20. The computer-implemented method of claim 19, further comprising: updating the predictive model based on the neural networks of which training is not terminated. 21. A computer program product comprising one or more computer readable mediums collectively including instructions that, when executed by the processor, cause the processor to: train a first neural network with a learning setting; extract tentative weight data from the first neural network with the learning setting; calculate an evaluation value of the first neural network with the learning setting; and generate a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network. 22. The computer program product of claim 21, wherein the calculating the evaluation value calculates the evaluation value of the first neural network with the learning setting and weight data further trained from the tentative weight data. 23. The computer program product of claim 22, wherein the instructions further cause the processor to: train the second neural network with the new setting; and estimate the evaluation value of the second neural network with the new setting by using the predictive model before completion of training the second neural network with the new setting. 24. The computer program product of claim 23, wherein the instructions further cause the processor to: terminate the training of the second neural network with the new setting in response to the evaluation value of the second neural network with the new setting not satisfying a criterion. 25. The computer program product of claim 24, wherein the instructions further cause the processor to: generate a plurality of new settings; train a plurality of neural networks, each neural network including a respective new setting among the plurality of new settings; terminate the training of at least one neural network among the plurality of neural networks that does not satisfy the criterion; and select one setting based on performances of neural networks of which training is not terminated. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Optimized learning settings of neural networks are efficiently determined by an apparatus including a processor and one or more computer readable mediums collectively including instructions that, when executed by the processor, cause the processor to train a first neural network with a learning setting; extract tentative weight data from the first neural network with the learning setting; calculate an evaluation value of the first neural network with the learning setting; and generate a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network. |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Optimized learning settings of neural networks are efficiently determined by an apparatus including a processor and one or more computer readable mediums collectively including instructions that, when executed by the processor, cause the processor to train a first neural network with a learning setting; extract tentative weight data from the first neural network with the learning setting; calculate an evaluation value of the first neural network with the learning setting; and generate a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network. |
|
Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor. | Please help me write a proper abstract based on the patent claims. CLAIM: 1-20. (canceled) 21. A hardware circuit for processing an average pooling neural network layer of a neural network, the hardware circuit being configured to process the average pooling neural network layer by performing operations comprising: obtaining, by a matrix multiplication circuit of the hardware circuit, an input tensor to the average pooling neural network layer, wherein the matrix multiplication circuit is configured to generate a tensor corresponding to an output of the average pooling neural network layer; and generating, by the matrix multiplication circuit of the hardware circuit, the tensor corresponding to the output of the average pooling neural network layer from the input tensor to the average pooling neural network layer, the generating comprising: performing, by the matrix multiplication circuit of the hardware circuit, a convolution of the input tensor and a kernel to generate a first tensor, wherein the kernel has a size equal to a size of a window of the average pooling neural network layer and is composed of elements that are each an identity matrix; performing, by the matrix multiplication circuit of the hardware circuit, element-wise multiplication of the first tensor and a first masking tensor to generate a second tensor, wherein each element of the first masking tensor is a resealing factor that is determined based on a number of elements of the input tensor averaged to generate a corresponding element of the tensor corresponding to the output of the average pooling neural network layer; and performing, by the matrix multiplication circuit of the hardware circuit, element-wise multiplication of the second tensor and a second masking tensor to generate the tensor corresponding to the output of the average pooling neural network layer, wherein each element of the second masking tensor is an inverse of a common denominator of the elements of the first masking tensor that are each a rescaling factor that is determined based on a number of elements of the input tensor averaged to generate a corresponding element of the tensor corresponding to the output of the average pooling neural network layer. 22. The hardware circuit of claim 21, wherein the input tensor is a zero-padded version of an initial input tensor, and wherein each element of the first masking tensor is a rescaling factor that is determined based on a number of elements of the initial input tensor that are averaged to generate a corresponding element of the tensor corresponding to the output of the average pooling neural network layer. 23. The hardware circuit of claim 22, wherein the zero-padded version of the initial input tensor is a version of the initial input tensor comprising, for each row and column of the initial input tensor, one or more preceding zeros and trailing zeros, wherein the number of preceding zeros and trailing zeros is determined based at least on the size of the window of the average pooling neural network layer. 24. The hardware circuit of claim 21, wherein a bit resolution of the elements of the first tensor is higher than a bit resolution of the elements of the second tensor. 25. The hardware circuit of claim 21, wherein a size of the first masking tensor and a size of the second masking tensor are each determined based at least on a size of the input tensor. 26. The hardware circuit of claim 21, wherein each of the first masking tensor and the second masking tensor comprise one or more masking tensor fragments that are tiled to generate a masking tensor. 27. The hardware circuit of claim 21, wherein the input tensor is stored at a unified buffer of the hardware circuit and the kernel is stored at a dynamic memory of the hardware circuit, and wherein the hardware circuit is configured to enable the matrix multiplication circuit of the hardware circuit to perform the convolution of the input tensor and the kernel to generate the first tensor by: sending the input tensor from the unified buffer to the matrix multiplication circuit of the hardware circuit; and sending the kernel from the dynamic memory to the matrix multiplication circuit of the hardware circuit. 28. The hardware circuit of claim 21, wherein each of the convolution of the input tensor and the kernel to generate the first tensor, the element-wise multiplication of the first tensor and the first masking tensor to generate the second tensor, and the element-wise multiplication of the second tensor and the second masking tensor to generate the tensor corresponding to the output of the average pooling neural network layer are performed as fixed point operations. 29. The hardware circuit of claim 21, wherein, to perform the convolution of the input tensor and the kernel to generate the first tensor, the matrix multiplication circuit of the hardware circuit is configured to: receive elements of the input tensor at one or more cells of the matrix multiplication circuit of the hardware circuit; receive weights of the kernel at the one or more cells of the matrix multiplication circuit of the hardware circuit; process the received elements of the input tensor and the received weights of the kernel at the one or more cells of the matrix multiplication circuit of the hardware circuit; and output results of the processing to one or more accumulators of the matrix multiplication circuit of the hardware circuit. 30. The hardware circuit of claim 29, wherein each cell of the matrix multiplication circuit of the hardware circuit comprises: an activation register configured to receive an element of the input tensor; a weight register configured to receive a weight of the kernel; and multiplication circuitry configured to multiply the element of the input tensor and the weight of the kernel. 31. The hardware circuit of claim 21, wherein, to perform the element-wise multiplication of the first tensor and the first masking tensor to generate the second tensor, the matrix multiplication circuit of the hardware circuit is configured to: perform matrix multiplication of the first tensor and an identity matrix to obtain a first output tensor, perform matrix multiplication of the first masking tensor and an identity matrix to obtain a second output tensor; and multiply the first output tensor and the second output tensor to obtain the second tensor. 32. (canceled) 33. The hardware circuit of claim 21, wherein, to perform the element-wise multiplication of the second tensor and the second masking tensor to generate the third tensor, the matrix multiplication circuit of the hardware circuit is configured to: perform matrix multiplication of the second tensor and an identity matrix to obtain a first output tensor, perform matrix multiplication of the second masking tensor and an identity matrix to obtain a second output tensor; and multiply the first output tensor and the second output tensor to obtain the third tensor. 34. (canceled) 35. A hardware circuit for processing an average pooling neural network layer of a neural network, the hardware circuit being configured to process the average pooling neural network layer by performing operations comprising: obtaining, by a matrix multiplication circuit of the hardware circuit, an input tensor to the average pooling neural network layer, wherein the matrix multiplication circuit is configured to generate a tensor corresponding to an output of the average pooling neural network layer; and generating, by the matrix multiplication circuit of the hardware circuit, the tensor corresponding to the output of the average pooling neural network layer from the input tensor to the average pooling neural network layer, the generating comprising: performing, by the matrix multiplication circuit of the hardware circuit, a convolution of the input tensor and a kernel to generate a first tensor, wherein the kernel has a size equal to a size of a window of the average pooling neural network layer and is composed of elements that are each an identity matrix; performing, by the matrix multiplication circuit of the hardware circuit, element-wise multiplication of the first tensor and a first masking tensor to generate a second tensor, wherein each element of the first masking tensor is a least common denominator of (i) a number of elements of the input tensor averaged to generate a corner element of the tensor corresponding to the output of the average pooling neural network layer, (ii) a number of elements of the input tensor averaged to generate an edge element of the tensor corresponding to the output of the average pooling neural network layer, and (iii) a number of elements in the kernel; and performing, by the matrix multiplication circuit of the hardware circuit, element-wise multiplication of the second tensor and a second masking tensor to generate the tensor corresponding to the output of the average pooling neural network layer, wherein each element of the second masking tensor is a resealing factor that is determined based on a number of elements of the input tensor that are averaged to generate a corresponding element of the tensor corresponding to the output of the average pooling neural network layer. 36. The hardware circuit of claim 35, wherein the input tensor is a zero-padded version of an initial input tensor, and wherein each element of the second masking tensor is a resealing factor that is determined based on a number of elements of the initial input tensor that are averaged to generate a corresponding element of the tensor corresponding to the output of the average pooling neural network layer. 37. The hardware circuit of claim 36, wherein the zero-padded version of the initial input tensor is a version of the initial input tensor comprising, for each row and column of the initial input tensor, one or more preceding zeros and trailing zeros, wherein the number of preceding zeros and trailing zeros is determined based at least on the size of the window of the average pooling neural network layer. 38. The hardware circuit of claim 35, wherein a size of the first masking tensor and a size of the second masking tensor are each determined based at least on a size of the input tensor. 39. The hardware circuit of claim 35, wherein each of the convolution of the input tensor and the kernel to generate the first tensor, the element-wise multiplication of the first tensor and the first masking tensor to generate the second tensor, and the element-wise multiplication of the second tensor and the second masking tensor to generate the tensor corresponding to the output of the average pooling neural network layer are performed as fixed point operations. 40. The hardware circuit of claim 35, wherein, to perform the convolution of the input tensor and the kernel to generate the first tensor, the matrix multiplication circuit of the hardware circuit is configured to: receive elements of the input tensor at one or more cells of the matrix multiplication circuit of the hardware circuit computation circuitry; receive weights of the kernel at the one or more cells of the matrix multiplication circuit of the hardware circuit; process the received elements of the input tensor and the received weights of the kernel at the one or more cells of the matrix multiplication circuit of the hardware circuit; and output results of the processing to one or more accumulators of the matrix multiplication circuit of the hardware circuit. 41. The hardware circuit of claim 21, wherein the matrix multiplication circuit of the hardware circuit is not configured to directly process the input tensor to generate the tensor corresponding to the output of the average pooling neural network layer. 42. The hardware circuit of claim 21, wherein a size of the first masking tensor and a size of the second masking tensor are each the same as a size of the input tensor. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor. |
|
G06N30635 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor. |
|
A recommender engine is configured to access memory and surface transmedia content items; and/or linked transmedia content subsets; and/or one or more identifications of identified users; and/or content items of the plurality of transmedia content items associated with at least one identified user. The surfaced items are presented for selection by the given user via the transmedia content linking engine as one or more user-selected transmedia content items. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An apparatus for surfacing transmedia content to a given user of a plurality of users comprising: a memory configured to store: a plurality of transmedia content items, each content item being associated with at least one of the plurality of users; and linking data which define time-ordered content links between the plurality of transmedia content items, the plurality of transmedia content items being arranged into linked transmedia content subsets comprising different groups of the transmedia content items and different content links therebetween; a transmedia content linking engine configured to receive user input indicative of a time-ordering between at least two user-selected transmedia content items and generate the content link data for storage in the memory, thereby defining a linked transmedia content subset including the at least two user-selected transmedia content items; and a recommender engine configured to access the memory and surface to the given user: one or more content items of the plurality of transmedia content items; and/or one or more of the linked transmedia content subsets of the linked transmedia content subsets; and/or one or more identifications of identified users other than the given user; and/or the content items of the plurality of transmedia content items associated with at least one identified user other than the given user, wherein the one or more surfaced content items and/or the one or more surfaced linked transmedia content subsets are surfaced for selection by the given user via the transmedia content linking engine as one or more user-selectable transmedia content items. 2. The apparatus of claim 1, wherein the memory further includes user ownership data which associates each transmedia content item to a user of the plurality of users, and each transmedia content subset of the multiple content subsets with a user of the plurality of users. 3. The apparatus of claim 2, wherein the transmedia content linking engine is configured to identify a current user-selected time-ordered location of the given user within a given linked transmedia content subset for insertion of one of the transmedia content items, and to generate corresponding content linking data therefor upon insertion by the given user of the one or more content items into the given linked transmedia content subset. 4. The apparatus of claim 3, wherein the recommender engine is further configured to surface the one or more transmedia content items and/or surface the one or more transmedia content subsets based on the current user-selected time-ordered location. 5. The apparatus of claim 1, further comprising a preference model, wherein the preference model is configured to provide a prediction of a rating for a given user for a given transmedia content data item. 6. The apparatus of claim 5, wherein the preference model is configured to: remove global effects for one or more rated items that have been previously rated by the given user to generate explicit ratings and storing the removed global effects; provide the explicit ratings and metadata associated with the given transmedia content data item to a plurality of recommender algorithms to generate a plurality of predicted ratings for the given transmedia content data item; combine the plurality of predicted ratings to build an ensemble prediction of the predicted ratings for the given transmedia content data item; add the stored global effects to the ensemble prediction to produce a predicted user rating for the given transmedia content data item; and output the predicted user rating for the given transmedia content data item. 7. The apparatus of claim 5, wherein the apparatus further comprises a user-brand match component configured to provide for a given user a prediction of a preference for a given branded content data item. 8. The apparatus of claim 7, wherein the apparatus further comprises a branded content model configured to provide for a given transmedia content item a prediction of the suitability of a given branded content data item. 9. The apparatus of claim 8, wherein the recommender engine is configured to surface transmedia content data items by querying the preference model, user-brand match component and brand model component by providing the preference model, user-brand match component and brand model with a transmedia content parameter, user data for the given user and a given branded content data item, and to maximise the sum of the output for the preference model, user-brand match component and brand model over the transmedia content parameter. 10. The apparatus of claim 9, wherein the recommender engine is configured to surface the transmedia content data item with the maximum output. 11. The apparatus of claim 1, wherein, the plurality of transmedia content items comprises items of different content types. 12. The apparatus of claim 1, wherein the transmedia content data items relate to narrative elements of the transmedia content data items. 13. The apparatus of claim 12, wherein the time-ordered content links define a narrative order of the transmedia content data items. 14. The apparatus of claim 1, wherein each time-ordered content link defines a directional link from a first transmedia content data item to a second transmedia content data item of the plurality of transmedia content data items. 15. The apparatus of claim 14, wherein the first transmedia content data item has a plurality of outgoing time-ordered content links. 16. The apparatus of claim 14, wherein the second transmedia content data item has a plurality of incoming time-ordered content links. 17. The apparatus of claim 14, wherein the memory is further configured to store a plurality of subset entry points for the plurality of transmedia content subsets. 18. The apparatus of claim 17, wherein each subset entry point is a flag indicating a transmedia content data item that has at least one outgoing time-ordered link and no incoming time-ordered links. 19. The apparatus of claim 18, wherein each linked transmedia content subset defines a linear path, wherein a linear path comprises a subset entry point, one or more transmedia content data items and one or more time ordered links between the subset entry point and the transmedia content data items. 20. The apparatus of claim 19, wherein two or more transmedia content subsets share one or more subset entry points, one or more transmedia content data items and/or one or more time ordered content links. 21. The apparatus of claim 21, wherein the recommender engine is further configured to surface one or more groups transmedia content subsets to the given user, the one or more surfaced groups being surfaced for selection by the given user via the transmedia content linking engine. 22. The apparatus of claim 1, further comprising a user model, wherein the user model is configured to provide a predictions of the state and/or behaviour of a given user to the recommender engine. 23. The apparatus of claim 22, wherein the memory is further configured to store a plurality of historical user interaction data items which define the historical interaction of a given user with the apparatus for surfacing transmedia content to a given user of a plurality of users, and wherein the user model is configured to: receive one or more historical user interaction data items associated with the given user; provide the one or more historical interaction data items to a statistical model to produce a prediction of a mental state of the user; retrieve from the statistical model the predicted mental state of the user and/or a predicted behaviour of the user; and output the predicted mental state of the user and/or predicted behaviour of the user. 24. The apparatus of claim 23, wherein the statistical model is a hidden Markov model. 25. The apparatus of claim 23, wherein the transmedia content recommender engine is further configured to surface the one or more individual content items to the given user and/or surface one or more of the linked transmedia content subsets to the given user when the output of the user model indicates the user is in mental state conducive to consuming transmedia content data items. 26. The apparatus of claim 23, wherein the transmedia content recommender engine is further configured to surface the one or more individual content items to the given user and/or surface one or more of the linked transmedia content subsets to the given user when the output of the user model indicates the predicted user behaviour is to consume transmedia content data items. 27. A system for surfacing transmedia content to a given user of a plurality of users comprising: the apparatus of claim 1; an electronic device configured to be in communication with the apparatus and receive and display to the given user: the surfaced one or more content items of the plurality of transmedia content items; and/or the surfaced one or more linked transmedia content subsets of the linked transmedia content subsets; and/or the surfaced one or more identifications of identified users other than the given user; and/or the surfaced content items of the plurality of transmedia content items associated with at least one identified user other than the given user. 28. A method for surfacing transmedia content to a given user of a plurality of users from a memory configured to store a plurality of transmedia content items, each content item being associated with at least one of the plurality of users, and linking data which define time-ordered content links between the plurality of transmedia content items, the plurality of transmedia content items being arranged into linked transmedia content subsets comprising different groups of the transmedia content items and different content links therebetween, the method comprising: receiving, at a transmedia content linking engine, user input indicative of a time-ordering between at least two user-selected transmedia content items; generating with the transmedia content linking engine the content linking data for storage in the memory, thereby defining a linked transmedia content subset including the at least two user-selected transmedia content items; and accessing the memory and surfacing to the given user with a recommender engine: one or more content items of the plurality of transmedia content items; and/or one or more of the linked transmedia content subsets of the linked transmedia content subsets; and/or one or more identifications of identified users other than the given user; and/or the content items of the plurality of transmedia content items associated with at least one identified user other than the given user, wherein the one or more surfaced content items and/or the one or more surfaced linked transmedia content subsets are surfaced for selection by the given user via the transmedia content linking engine as one or more user-selectable transmedia content items. 29. A computer readable medium comprising computer executable instructions, which when executed by a computer, cause the computer to perform the steps of the method of claim 28. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A recommender engine is configured to access memory and surface transmedia content items; and/or linked transmedia content subsets; and/or one or more identifications of identified users; and/or content items of the plurality of transmedia content items associated with at least one identified user. The surfaced items are presented for selection by the given user via the transmedia content linking engine as one or more user-selected transmedia content items. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A recommender engine is configured to access memory and surface transmedia content items; and/or linked transmedia content subsets; and/or one or more identifications of identified users; and/or content items of the plurality of transmedia content items associated with at least one identified user. The surfaced items are presented for selection by the given user via the transmedia content linking engine as one or more user-selected transmedia content items. |
|
A computer device for generating a classifier for performing a query to a given knowledge base is provided. The given knowledge base includes predicates, subjects and objects related to each other. The computer device includes a selection entity for selecting one of the predicates, and a triple generation entity for generating, based on the given knowledge base, triples. Each of the triples includes the one selected predicate, and a subject and an object related to the one selected predicate. The computer device also includes a candidate generation entity for generating a list of properties. Each property of the list of properties is correlated to the subject and the object of one of the triples by performing a context-based query within the given knowledge base. The computer device includes a classifier generation entity for generating a classifier having the list of properties related to the selected predicate. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer device for generating a classifier for performing a query to a given knowledge base, the given knowledge base including a plurality of predicates, a plurality of subjects and a plurality objects related to each other, the computer device comprising: a processor configured to: select one predicate from the plurality of predicates; generate, based on the given knowledge base, a plurality of triples, each triple of the plurality of triples including the one selected predicate, and a subject of the plurality of subjects and an object of the plurality of objects related to the one selected predicate; generate a list of properties, each property of the list of properties being correlated to the subject and the object of the triple of the plurality of triples, the generation of the list of properties comprising performance of a context-based query within the given knowledge base; and generate a classifier having the list of properties related to the one selected predicate. 2. The computer device of claim 1, wherein the processor is further configured to select subjects of the plurality of subjects and objects of the plurality of objects each having the one selected predicate. 3. The computer device of claim 1, wherein the plurality of triples are RDF-based triples, and wherein the subjects and objects of the plurality of triples are URI-based subjects and URI-based objects, respectively. 4. The computer device of claim 1, wherein the processor is further configured to process a reference text of the given knowledge base to find at least one sentence with an occurrence of the subject and the object of one triple of the plurality of triples. 5. The computer device of claim 4, wherein the processor is further configured, when processing the reference text, to generate a search query, the search query requesting sentences, in which the subject and the object of the one triple of the plurality of triples occur. 6. The computer device of claim 4, wherein the processor is further configured to process the reference text for each triple of the plurality of triples. 7. The computer device of claim 4, wherein the processor is further configured to convert subjects and objects occurring in requested sentences into generic subjects and generic objects. 8. The computer device of claim 7, wherein the processor is further configured to generate the classifier based on a matrix, in which each of the requested sentences is represented as a feature vector. 9. The computer device of claim 1, wherein the processor is further configured to: select a number of predicates from the plurality of predicates; generate a number of triples for each predicate of the number of predicates; generate a list of properties for each of the number of triples; and generate a classifier model including a plurality of classifiers having the list of properties related to each of the number of selected predicates. 10. The computer device of claim 9, wherein the processor is further configured to generate the classifier model, the generation of the classifier model comprising use of a support vector machine. 11. The computer device of claim 10, wherein the processor is further configured to use the support vector machine with the generated matrix. 12. The computer device of claim 1, wherein the processor is further configured to receive a natural language query and to process the natural language query using the generated classifier. 13. The computer device of claim 12, wherein the processor is further configured to process the natural language query using the generated classifier by converting the natural language query into a feature vector using the generated matrix. 14. A method for generating a classifier for performing a query to a given knowledge base, the given knowledge base including a plurality of predicates, a plurality of subjects and a plurality objects being related to each other, the method comprising: selecting one predicate from the plurality of predicates; generating, based on the given knowledge base, a plurality of triples, each triple of the plurality of triples including the one selected predicate and a subject of the plurality of subjects and an object of the plurality of objects related to the one selected predicate; generating a list of properties, the generating of the list of properties comprising performing a context-based query within the given knowledge base, each property of the list of properties being correlated to the subject and the object of one triple of the plurality of triples; and generating a classifier having the list of properties related to the one selected predicate. 15. The method of claim 14, further comprising receiving a natural language query and processing the natural language query using the generated classifier. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: A computer device for generating a classifier for performing a query to a given knowledge base is provided. The given knowledge base includes predicates, subjects and objects related to each other. The computer device includes a selection entity for selecting one of the predicates, and a triple generation entity for generating, based on the given knowledge base, triples. Each of the triples includes the one selected predicate, and a subject and an object related to the one selected predicate. The computer device also includes a candidate generation entity for generating a list of properties. Each property of the list of properties is correlated to the subject and the object of one of the triples by performing a context-based query within the given knowledge base. The computer device includes a classifier generation entity for generating a classifier having the list of properties related to the selected predicate. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A computer device for generating a classifier for performing a query to a given knowledge base is provided. The given knowledge base includes predicates, subjects and objects related to each other. The computer device includes a selection entity for selecting one of the predicates, and a triple generation entity for generating, based on the given knowledge base, triples. Each of the triples includes the one selected predicate, and a subject and an object related to the one selected predicate. The computer device also includes a candidate generation entity for generating a list of properties. Each property of the list of properties is correlated to the subject and the object of one of the triples by performing a context-based query within the given knowledge base. The computer device includes a classifier generation entity for generating a classifier having the list of properties related to the selected predicate. |
|
A machine learning apparatus, which learns a condition associated with a filter unit for filtering an analog input signal, includes a state observer for observing a state variable that includes at least one of a noise component and noise amount of an output signal from the filter unit and a responsivity to the input signal; and a learner for learning the condition associated with the filter unit in accordance with a training data set that includes the state variable. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A machine learning apparatus for learning a condition associated with a filter unit for filtering an analog input signal, comprising: a state observer for observing a state variable that includes a noise component and noise amount of an output signal from the filter unit and a responsivity to the input signal; and a learner for learning the condition associated with the filter unit in accordance with a training data set that includes the state variable. 2. The machine learning apparatus according to claim 1, wherein the state observer further observes information about a machine state from a machine commander; and the learner further updates the training data set based on the information about the machine state. 3. The machine learning apparatus according to claim 1, wherein the learner learns the condition in accordance with training data sets acquired on a plurality of filter units. 4. The machine learning apparatus according to claim 1, further comprising a decision maker for updating a filter based on a result of learning by the learner in accordance with the training data set. 5. The machine learning apparatus according to claim 1, further comprising a reward calculator for calculating a reward based on the noise component, the noise amount, and the responsivity; and a function updater for updating a function to modify the filter unit from the current state variable based on the reward. 6. The machine learning apparatus according to claim 5, wherein the reward calculator decreases the reward when the noise amount is increased or the responsivity is lower than a specified value, while the reward calculator increases the reward when the noise amount is decreased and the responsivity is equal to or higher than the specified value. 7. The machine learning apparatus according to claim 5, wherein the learner relearns and updates the condition in accordance with an additional training data set that includes the current state variable. 8. The machine learning apparatus according to claim 7, wherein the machine learning apparatus is connected to the filter unit through a network; and the state observer acquires the current state variable through the network. 9. A motor drive apparatus comprising: the machine learning apparatus according to claim 1; and a filter unit having a variable filter and a filter modifier for modifying the variable filter. 10. The motor drive apparatus according to claim 9, wherein the machine learning apparatus is present in a cloud server. 11. A motor drive system comprising: the motor drive apparatus according to claim 9; and a machine commander for providing notification of information about the operation of a machine. 12. A machine learning method for learning a condition associated with a filter unit for filtering an analog input signal, comprising the steps of: observing a state variable that includes a noise component and noise amount of an output signal from the filter unit and a responsivity to the input signal; and learning the condition associated with the filter unit in accordance with a training data set that includes the state variable. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A machine learning apparatus, which learns a condition associated with a filter unit for filtering an analog input signal, includes a state observer for observing a state variable that includes at least one of a noise component and noise amount of an output signal from the filter unit and a responsivity to the input signal; and a learner for learning the condition associated with the filter unit in accordance with a training data set that includes the state variable. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A machine learning apparatus, which learns a condition associated with a filter unit for filtering an analog input signal, includes a state observer for observing a state variable that includes at least one of a noise component and noise amount of an output signal from the filter unit and a responsivity to the input signal; and a learner for learning the condition associated with the filter unit in accordance with a training data set that includes the state variable. |
|
The present disclosure provides a human-computer intelligence chatting method and device. The method includes: receiving a multimodal input signal, the multimodal input signal comprising at least one of a speech signal, an image signal, a sensor signal and an event driving signal; processing the multimodal input signal to obtain text data, and obtaining an intention of a user according to the text data; obtaining an answer corresponding to the intention of the user, and converting the answer to a multimodal output signal; and outputting the multimodal output signal. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A human-computer intelligence chatting method based on artificial intelligence, comprising: receiving a multimodal input signal, the multimodal input signal comprising at least one of a speech signal, an image signal, a sensor signal and an event driving signal; processing the multimodal input signal to obtain text data, and obtaining an intention of a user according to the text data; obtaining an answer corresponding to the intention of the user, and converting the answer to a multimodal output signal; and outputting the multimodal output signal. 2. The method according to claim 1, wherein obtaining an intention of a user according to the text data comprises: analyzing the text data and generating the intention of the user according to a result of analyzing the text data. 3. The method according to claim 2, wherein analyzing the text data comprises: performing a syntactic structure analysis on the text data, and performing a semantic analysis based on words, a domain multi-classification recognition based on a topic model, a semantic disambiguation, and an auto-completion based on grammatical structures and context information. 4. The method according to claim 1, further comprising: storing the intention of the user into historical intentions of the user. 5. The method according to claim 1, wherein obtaining an answer corresponding to the intention of the user comprises: searching a memory system according to the intention of the user, so as to obtain constraint conditions on the intention of the user; searching a topic model and a domain entity database according to the intention of the user, so as to obtain variables and attributes associated with the intention of the user; obtaining a similarity between a current chat context and a pre-stored chat mode via an active learning module; accessing an open service interface, and obtaining a result returned via the open service interface; and obtaining the answer corresponding to the intention of the user according to the intention of the user in combination with the constraint conditions on the intention of the user, the variables and attributes associated with the intention of the user, the result returned via the open service interface and the similarity between the current chat context and the pre-stored chat mode. 6. The method according to claim 5, further comprising: storing the intention of the user, the constraint conditions on the intention of the user, and the variables and attributes associated with the intention of the user into a dialogue model; establishing a transition probability map according to a statistical result stored in the dialogue model, and generating a new topic according to the transition probability map at an appropriate time, wherein the statistical result is obtained according to the intention of the user, the constraint conditions on the intention of the user, and the variables and attributes associated with the intention of the user. 7. The method according to claim 5, after processing the multimodal input signal to obtain text data, further comprising: storing content favorable for memorizing into the memory system after obtaining the text data. 8. The method according to claim 7, wherein the memory system comprises a short-term memory system and a long-term memory system; storing content favorable for memorizing into the memory system comprises: storing content favorable for short-term memorizing into the short-term memory system, in which the content favorable for short-term memorizing comprises historical dialogue records of the user, a topic status sequence established based on the historical dialogue records and entity-related attributes extracted from the historical dialogue records; and storing content favorable for long-term memorizing into the long-term memory system, in which the content favorable for long-term memorizing comprises personal information and population attributes of the user, preferences of the user, historical geographic records of the user, historical purchase records of the user, personal information and population attributes in the system, and preferences of the system. 9. The method according to claim 5, after processing the multimodal input signal to obtain text data, further comprising: recording topics extracted from the text data in the topic model and recording entity attributes extracted from the text data in the domain entity database. 10. The method according to claim 5, wherein obtaining a similarity between a current chat context and a pre-stored chat mode via an active learning module comprises: performing a numeralization on the chat mode of human beings according to the dialogue model, the topic model and the domain entity database, so as to obtain a numerical chat mode; storing the numerical chat mode in the active learning module; and detecting by the active learning module, the similarity between the current chat context and the numerical chat mode. 11. A human-computer intelligence chatting device based on artificial intelligence, comprising: a processor; and a memory, configured to store instructions executable by the processor, wherein, the processor is configured to: receive a multimodal input signal, the multimodal input signal comprising at least one of a speech signal, an image signal, a sensor signal and an event driving signal; process the multimodal input signal to obtain text data, and obtain an intention of a user according to the text data; obtain an answer corresponding to the intention of the user, and convert the answer to a multimodal output signal; and output the multimodal output signal. 12. The device according to claim 11, wherein the processor is configured to analyze the text data and generate the intention of the user according to a result of analyzing the text data, in which the processor is specifically configured to perform a syntactic structure analysis on the text data, and perform a semantic analysis based on words, a domain multi-classification recognition based on a topic model, a semantic disambiguation, and an auto-completion based on grammatical structures and context information. 13. The device according to claim 11, wherein the processor is further configured to store the intention of the user into historical intentions of the user. 14. The device according to claim 11, wherein the processor is configured to: search a memory system according to the intention of the user, so as to obtain constraint conditions on the intention of the user; search a topic model and a domain entity database according to the intention of the user, so as to obtain variables and attributes associated with the intention of the user; obtain a similarity between a current chat context and a pre-stored chat mode via the active learning module; access an open service interface and obtain a result returned via the open service interface; and obtain the answer corresponding to the intention of the user according to the intention of the user in combination with the constraint conditions on the intention of the user, the variables and attributes associated with the intention of the user, the result returned via the open service interface and the similarity between the current chat context and the pre-stored chat mode. 15. The device according to claim 14, wherein the processor is further configured to: store the intention of the user, the constraint conditions on the intention of the user, and the variables and attributes associated with the intention of the user into a dialogue model; establish a transition probability map according to a statistical result stored in the dialogue model, and generate a new topic according to the transition probability map at an appropriate time, wherein the statistical result is obtained according to the intention of the user, the constraint conditions on the intention of the user, and the variables and attributes associated with the intention of the user. 16. The device according to claim 14, wherein the processor is further configured to: store content favorable for memorizing into the memory system after obtaining the text data. 17. The device according to claim 16, wherein the memory system comprises a short-term memory system and a long-term memory system, and the processor is further configured to store content favorable for short-term memorizing into the short-term memory system, and to store content favorable for long-term memorizing into the long-term memory system, in which the content favorable for short-term memorizing comprises historical dialogue records of the user, a topic status sequence established based on the historical dialogue records and entity-related attributes extracted from the historical dialogue records; and the content favorable for long-term memorizing comprises personal information and population attributes of the user, preferences of the user, historical geographic records of the user, historical purchase records of the user, personal information and population attributes in the system and preferences of the system. 18. The device according to claim 14, wherein the processor is further configured to: record topics extracted from the text data in the topic model, and record entity attributes extracted from the text data in the domain entity database after obtaining the text data. 19. The device according to claim 14, wherein the processor is configured to: perform a numeralization on the chat mode of human beings according to the dialogue model, the topic model and the domain entity database, so as to obtain a numerical chat mode; store the numerical chat mode in the active learning module; and obtain the similarity between the current chat context and the numerical chat mode from the active learning module. 20. A non-transitory computer-readable storage medium having stored therein instructions that, when executed by a processor of a device, causes the device to perform a human-computer intelligence chatting method based on artificial intelligence, the method comprising: receiving a multimodal input signal, the multimodal input signal comprising at least one of a speech signal, an image signal, a sensor signal and an event driving signal; processing the multimodal input signal to obtain text data, and obtaining an intention of a user according to the text data; obtaining an answer corresponding to the intention of the user, and converting the answer to a multimodal output signal; and outputting the multimodal output signal. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: The present disclosure provides a human-computer intelligence chatting method and device. The method includes: receiving a multimodal input signal, the multimodal input signal comprising at least one of a speech signal, an image signal, a sensor signal and an event driving signal; processing the multimodal input signal to obtain text data, and obtaining an intention of a user according to the text data; obtaining an answer corresponding to the intention of the user, and converting the answer to a multimodal output signal; and outputting the multimodal output signal. |
|
G06N3006 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: The present disclosure provides a human-computer intelligence chatting method and device. The method includes: receiving a multimodal input signal, the multimodal input signal comprising at least one of a speech signal, an image signal, a sensor signal and an event driving signal; processing the multimodal input signal to obtain text data, and obtaining an intention of a user according to the text data; obtaining an answer corresponding to the intention of the user, and converting the answer to a multimodal output signal; and outputting the multimodal output signal. |
|
Disclosed herein is a deep learning model that can be used for performing speech or image processing tasks. The model uses multi-task training, where the model is trained for at least two inter-related tasks. For face detection, the first task is face detection (i.e. face or non-face) and the second task is facial feature identification (i.e. mouth, eyes, nose). The multi-task model improves the accuracy of the task over single-task models. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for performing speech or image processing tasks comprising: training a deep learning model with at least two inter-related tasks; and processing at least one of an image or an audio clip using the deep learning model. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Disclosed herein is a deep learning model that can be used for performing speech or image processing tasks. The model uses multi-task training, where the model is trained for at least two inter-related tasks. For face detection, the first task is face detection (i.e. face or non-face) and the second task is facial feature identification (i.e. mouth, eyes, nose). The multi-task model improves the accuracy of the task over single-task models. |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Disclosed herein is a deep learning model that can be used for performing speech or image processing tasks. The model uses multi-task training, where the model is trained for at least two inter-related tasks. For face detection, the first task is face detection (i.e. face or non-face) and the second task is facial feature identification (i.e. mouth, eyes, nose). The multi-task model improves the accuracy of the task over single-task models. |
|
A convolution operation method includes the following steps of: decomposing a large convolution operation region to multiple small convolution operation regions; the small convolution operation regions perform convolution operations so as to generate partial results, respectively; and summing the partial results as a convolution operation result of the large convolution operation region. A convolution operation device capable of supporting the convolution operation method is also disclosed. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A convolution operation method, comprising following steps of: decomposing a large convolution operation region to multiple small convolution operation regions; performing convolution operations by the small convolution operation regions so as to generate partial results, respectively; and summing the partial results as a convolution operation result of the large convolution operation region. 2. The convolution operation method of claim 1, wherein the small convolution operation regions have the same scale. 3. The convolution operation method of claim 1, further comprising a step of: assigning 0 to the small convolution operation regions, which are exceeding the large convolution operation region. 4. The convolution operation method of claim 1, wherein, in the step of performing the convolution operations, the small convolution operation regions utilize at least a convolution unit to perform the convolution operations so as to generate the partial results, and a scale of the small convolution operation region is equal to a maximum convolution scale capable of being supported by the convolution unit. 5. The convolution operation method of claim 1, wherein, in the step of performing the convolution operations, the small convolution operation regions utilize convolution units of corresponding numbers to perform the convolution operations in parallel so as to generate the partial results. 6. The convolution operation method of claim 1, wherein the large convolution operation region comprises a plurality of filter coefficients, and the filter coefficients are assigned to the small convolution operation regions according to an order of the filter coefficients and scales of the small convolution operations regions. 7. The convolution operation method of claim 1, wherein the large convolution operation region comprises a plurality of data, and the filter coefficients are assigned to the small convolution operation regions according to an order of the data and scales of the small convolution operations regions. 8. The convolution operation method of claim 1, wherein a scale of the large convolution operation region is 5×5 or 7×7, and a scale of the small convolution operation regions is 3×3. 9. The convolution operation method of claim 1, wherein the step of summing the partial results further comprises: providing a plurality of moving addresses to the small convolution operation regions, wherein the partial results move in a coordinate according to the moving addresses and added. 10. The convolution operation method of claim 1, further comprising: determining a convolution operation mode according to a scale of a current convolution operation region; wherein when the convolution operation mode is a decomposed mode, the current convolution operation region is the large convolution operation region, wherein the large convolution operation region is decomposed to the multiple small convolution operation regions, the small convolution operation regions perform the convolution operations so as to generate the partial results, respectively, and the partial results are summed as the convolution operation result of the large convolution operation region; and wherein when the convolution operation mode is a non-decomposed mode, the current convolution operation region is not decomposed and directly performs the convolution operation. 11. The convolution operation method of claim 1, further comprising: performing a partial operation of a consecutive layer of a convolutional neural network. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A convolution operation method includes the following steps of: decomposing a large convolution operation region to multiple small convolution operation regions; the small convolution operation regions perform convolution operations so as to generate partial results, respectively; and summing the partial results as a convolution operation result of the large convolution operation region. A convolution operation device capable of supporting the convolution operation method is also disclosed. |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A convolution operation method includes the following steps of: decomposing a large convolution operation region to multiple small convolution operation regions; the small convolution operation regions perform convolution operations so as to generate partial results, respectively; and summing the partial results as a convolution operation result of the large convolution operation region. A convolution operation device capable of supporting the convolution operation method is also disclosed. |
|
A bias estimation apparatus according to an embodiment estimates a bias included in a measured values by each sensor. The bias estimation apparatus includes a reference model builder, a temporary bias generator, a corrected measured value calculator, a similarity calculator, a similarity selector, a score calculator, and an estimated bias determiner. The reference model builder builds a reference model of the measured value packs. The temporary bias generator generates a temporary bias pack. The corrected measured value calculator calculates corrected measured value packs. The similarity calculator calculates a similarity of each corrected measured value pack. The similarity selector selects a part of the similarities according to their values from among the similarities. The score calculator calculates a score based on the selected similarities. The estimated bias determiner determines an estimated bias which is an estimated value of the bias based on the score. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A bias estimation apparatus for estimating a bias included in a measured values by each sensor, comprising: a reference model builder configured to build a reference model of the measured value packs each of which is measured values by sensors in a single measurement action; a temporary bias generator configured to generate a temporary bias pack which is temporary estimated values of the biases of sensors; a corrected measured value calculator configured to calculate corrected measured value packs each of which is the measured value pack corrected by the temporary bias pack; a similarity calculator configured to calculate a similarity of each corrected measured value pack relative to the reference model; a similarity selector configured to select a part of the similarities according to their values from among the similarities calculated by the similarity calculator; a score calculator configured to calculate a score which indicates a similarity of the whole corrected measured value packs relative to the reference model based on the selected similarities; and an estimated bias determiner configured to determine an estimated bias which is an estimated value of the bias for each sensor based on the score. 2. The apparatus according to claim 1, wherein the similarity selector selects the similarities in an order from the largest value. 3. The apparatus according to claim 1, wherein the score calculator calculates any one of an average value, a median, and a mode of the selected similarities as the score. 4. The apparatus according to claim 1, wherein the temporary bias generator generates the plurality of temporary bias packs, and the estimated bias determiner determines the temporary bias pack with the maximum score as the estimated bias pack of which each element is the estimated bias of a sensor from among the plurality of temporary bias packs. 5. The apparatus according to claim 1, wherein the reference model builder builds the reference model by using principal component analysis. 6. The apparatus according to claim 1, wherein the reference model builder builds the reference model by using a neural network. 7. The apparatus according to claim 1, wherein the similarity selector selects the similarity with the value which is the maximum value and the similarity with the value which is the median. 8. The apparatus according to claim 7, wherein the score calculator calculates a difference of square roots of the selected similarities as the score. 9. The apparatus according to claim 1, further comprising: a data converter configured to convert qualitative variables into quantitative variables, wherein the measured value packs include the quantitative variables converted from the qualitative variables by the data converter. 10. A bias estimation method for estimating a bias included in measured values by each sensor, comprising: building a reference model of the measured value packs each of which is measured values by sensors in a single measurement action; generating a temporary bias pack which is temporary estimated values of the biases of sensors; calculating corrected measured value packs each of which is the measured value pack corrected by the temporary bias pack; calculating a similarity of each corrected measured value pack relative to the reference model; selecting a part of the similarities according to their values from among the similarities calculated by the similarity calculator; calculating a score which indicates a similarity of the whole corrected measured value packs relative to the reference model based on the selected similarities; and determining an estimated bias which is an estimated value of the bias for each sensor based on the score. 11. The method according to claim 10, wherein the similarities are selected in an order from the largest value in selecting the similarities. 12. The method according to claim 10, further comprising: converting qualitative variables into quantitative variables, wherein the measured value packs include the quantitative variables converted from the qualitative variables. 13. A failure diagnosis apparatus for diagnosing failures of sensors, comprising: the apparatus according to claim 1; and a score calculator configured to calculate a score which indicates a similarity of the whole corrected measured value packs relative to the reference model based on the selected similarities; and a failure diagnoser configured to diagnose failures of sensors based on the score. 14. The apparatus according to claim 13, further comprising: a temporary bias generator configured to generate a temporary bias pack which is temporary estimated values of the biases of sensors; and a corrected measured value calculator configured to calculate corrected measured value packs each of which is the measured value pack corrected by the temporary bias pack, wherein the similarity calculator calculates a similarity of each corrected measured value pack relative to the reference model, the similarity selector selects a part of the similarities according to their values from among the similarities calculated by the similarity calculator, the score calculator calculates a score which indicates a similarity of the whole corrected measured value packs relative to the reference model based on the selected similarities, the estimated bias determiner determines an estimated bias which is an estimated value of the bias for each sensor based on the score, and the failure diagnoser diagnoses failures of the sensors based on the estimated biases. 15. A failure diagnosis method for diagnosing failures of the sensors, comprising: the method according to claim 10; and diagnosing failures of the sensors based on the score. 16. The method according to claim 15, further comprising: generating a temporary bias pack which is temporary estimated values of the biases of sensors; and calculating corrected measured value packs each of which is the measured value pack corrected by the temporary bias pack, wherein a similarity of each corrected measured value pack relative to the reference model is calculated, a part of the calculated similarities are selected according to their values, a score which indicates a similarity of the whole corrected measured value packs relative to the reference model is calculated based on the selected similarities, an estimated bias which is an estimated value of the bias is determined for each sensor based on the score, and failures of the sensors are diagnosed based on the estimated biases. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A bias estimation apparatus according to an embodiment estimates a bias included in a measured values by each sensor. The bias estimation apparatus includes a reference model builder, a temporary bias generator, a corrected measured value calculator, a similarity calculator, a similarity selector, a score calculator, and an estimated bias determiner. The reference model builder builds a reference model of the measured value packs. The temporary bias generator generates a temporary bias pack. The corrected measured value calculator calculates corrected measured value packs. The similarity calculator calculates a similarity of each corrected measured value pack. The similarity selector selects a part of the similarities according to their values from among the similarities. The score calculator calculates a score based on the selected similarities. The estimated bias determiner determines an estimated bias which is an estimated value of the bias based on the score. |
|
G06N5048 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A bias estimation apparatus according to an embodiment estimates a bias included in a measured values by each sensor. The bias estimation apparatus includes a reference model builder, a temporary bias generator, a corrected measured value calculator, a similarity calculator, a similarity selector, a score calculator, and an estimated bias determiner. The reference model builder builds a reference model of the measured value packs. The temporary bias generator generates a temporary bias pack. The corrected measured value calculator calculates corrected measured value packs. The similarity calculator calculates a similarity of each corrected measured value pack. The similarity selector selects a part of the similarities according to their values from among the similarities. The score calculator calculates a score based on the selected similarities. The estimated bias determiner determines an estimated bias which is an estimated value of the bias based on the score. |
|
A plurality of questions may be obtained for an AMA session. A priority for each of the plurality of questions may be calculated and a question to be answered from the plurality of questions may be determined based on question priority associated with each of the plurality of questions. An expert associated with the AMA session may be presented with the question. An answer to the question may be received from the expert associated with the AMA session and showing the answer to AMA participants may be facilitated. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A processor-implemented method to provide answers, comprising: obtaining via a processor a plurality of questions for an AMA session; calculating via the processor a priority for each of the plurality of questions; determining via the processor a question to be answered from the plurality of questions based on question priority associated with each of the plurality of questions; presenting via the processor an expert associated with the AMA session with the question; receiving via the processor an answer to the question from the expert associated with the AMA session; and facilitating via the processor showing the answer to AMA participants. 2. The method of claim 1, wherein the priority for each of the plurality of questions is determined based on the number of user votes for the corresponding question. 3. The method of claim 2, wherein each of the user votes is weighted when determining the priority. 4. The method of claim 1, wherein a plurality of experts is associated with the AMA session, and the expert presented with the question is chosen by a user who asked the question. 5. The method of claim 1, wherein a plurality of experts is associated with the AMA session, and the best-rated expert is chosen from the plurality of experts as the expert presented with the question. 6. The method of claim 1, wherein the answer is a video recording. 7. The method of claim 6, wherein the video recording is made via a plug-in of the expert's web browser. 8. The method of claim 6, further comprising: facilitating previewing the video recording by the expert; and obtaining another video recording of the answer. 9. The method of claim 6, wherein receiving the answer further comprises: receiving a selection from the expert of an audio background template; and adding the audio background template to the video recording. 10. The method of claim 6, wherein receiving the answer further comprises: identifying media related to the answer; obtaining a selection from the expert of media from the identified media; and embedding the selected media into the video recording. 11. The method of claim 1, further comprising facilitating alerting the AMA participants that the answer was received. 12. The method of claim 1, wherein facilitating showing the answer further comprises facilitating automatic playback of the answer for at least some of the AMA participants upon receiving the answer from the expert. 13. The method of claim 12, further comprising refraining from facilitating automatic playback of the subsequent answer received during the AMA session for an AMA participant while the AMA participant is commenting on the answer. 14. The method of claim 1, wherein facilitating showing the answer further comprises facilitating playback of answers received during the AMA session such that the answer is played back for at least some of the AMA participants in the order the answer was received from the expert in relation to other answers received during the AMA session. 15. The method of claim 14, further comprising facilitating skipping playback of a specified answer for an AMA participant if the AMA participant previously viewed the specified answer. 16. The method of claim 1, wherein facilitating showing the answer further comprises facilitating playback of answers received during the AMA session in the order specified by an AMA participant. 17. The method of claim 1, further comprising facilitating discussion of the answer by the AMA participants using a discussion component separate from discussion components associated with other answers received during the AMA session. 18. The method of claim 1, further comprising facilitating purchasing an item associated with the AMA session. 19. An answer providing apparatus, comprising: a memory; a processor in communication with the memory, and configured to issue a plurality of processing instructions stored in the memory, wherein the processor issues instructions to: obtain a plurality of questions for an AMA session; calculate a priority for each of the plurality of questions; determine a question to be answered from the plurality of questions based on question priority associated with each of the plurality of questions; present an expert associated with the AMA session with the question; receive an answer to the question from the expert associated with the AMA session; and facilitate showing the answer to AMA participants. 20. An answer providing processor-readable non-transitory physical medium storing processor-issuable instructions to: obtain a plurality of questions for an AMA session; calculate a priority for each of the plurality of questions; determine a question to be answered from the plurality of questions based on question priority associated with each of the plurality of questions; present an expert associated with the AMA session with the question; receive an answer to the question from the expert associated with the AMA session; and facilitate showing the answer to AMA participants. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: A plurality of questions may be obtained for an AMA session. A priority for each of the plurality of questions may be calculated and a question to be answered from the plurality of questions may be determined based on question priority associated with each of the plurality of questions. An expert associated with the AMA session may be presented with the question. An answer to the question may be received from the expert associated with the AMA session and showing the answer to AMA participants may be facilitated. |
|
G06N5022 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A plurality of questions may be obtained for an AMA session. A priority for each of the plurality of questions may be calculated and a question to be answered from the plurality of questions may be determined based on question priority associated with each of the plurality of questions. An expert associated with the AMA session may be presented with the question. An answer to the question may be received from the expert associated with the AMA session and showing the answer to AMA participants may be facilitated. |
|
This invention describes a system for utilizing dimensionalized archetypical or proxy representations of a person, place, thing, concept or construct and generally, a method for utilizing such representations for the purposes of information retrieval, knowledge management, and machine learning whereby the representations contribute to enhanced speed, contextual acuity, and overall value of the information stored within the system, as well as the easy utilization of the archetypical or proxy representations by such means or methods as weighted sorts, support vector machines, probabilistic filters, or other means whereby one or more of the dimensionalized tags or features represented by the affinitomic elements are utilized to make a selection. | Please help me write a proper abstract based on the patent claims. CLAIM: 1: What is claimed is a method and system for comparing a representational proxy for a real person, place, thing, concept, or construct within a computer system where such proxy is used to store tag elements for measuring or inferring affinity/nearness, or likelihood, wherein the system is comprised of a means of assigning a computer representation of a person, place, thing, concept or construct to a representation, archetype or proxy of said person, place, thing, concept or construct; a means of storing said representation such that it can be either securely and or publicly accessed by any such system that employs or seeks to employ such proxies or archetypes; a means of retrieving such proxies and archetypes which are deemed related to a specific feature or set of features within the proxy, archetype, or material represented by the proxy or archetype; The method and system of claim 1, further comprised of a means to construct a variety of kernel matrices from the elements and or features. 1. The method and system of claim 1, further comprised of a means whereby the archetypes or proxies are indexed or cached to affect rapid retrieval of those archetypes or proxies deemed related. 2. The method and system of claim 1, further comprised of a machine learning element that retrieves, evaluates, enhances, and or changes, improves or replaces the original archetype or proxy within the data store. 3. The method and system of claim 1 wherein the proxies and or archetypes and or constructs, such as kernels, constructed from these archetypes are, themselves, utilized as features to construct a proxy or archetype. 4. The method and system of claim 1 wherein the elements of the system reside across multiple systems that communicate or evaluate proxy or archetypical representations. 5. The method and system of claim 1 where the proxy representations are affinitomic archetypes. |
|
REJECTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: This invention describes a system for utilizing dimensionalized archetypical or proxy representations of a person, place, thing, concept or construct and generally, a method for utilizing such representations for the purposes of information retrieval, knowledge management, and machine learning whereby the representations contribute to enhanced speed, contextual acuity, and overall value of the information stored within the system, as well as the easy utilization of the archetypical or proxy representations by such means or methods as weighted sorts, support vector machines, probabilistic filters, or other means whereby one or more of the dimensionalized tags or features represented by the affinitomic elements are utilized to make a selection. |
|
G06N5048 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: This invention describes a system for utilizing dimensionalized archetypical or proxy representations of a person, place, thing, concept or construct and generally, a method for utilizing such representations for the purposes of information retrieval, knowledge management, and machine learning whereby the representations contribute to enhanced speed, contextual acuity, and overall value of the information stored within the system, as well as the easy utilization of the archetypical or proxy representations by such means or methods as weighted sorts, support vector machines, probabilistic filters, or other means whereby one or more of the dimensionalized tags or features represented by the affinitomic elements are utilized to make a selection. |
|
A question database storing questions and a conditional probability of one question to be asked given that a previous question was asked is searched to predict a future question based on the conditional probability stored in the question database given an input question as the previous question. The future question is suggested to a user. Responsive to receiving an acceptance of the future question, the question database is updated to strengthen the conditional probability associated with the future question occurring given the input question. An answer to the future question is provided and searching, predicting, suggesting and updating may be repeated, with the future question as the input question, until the future question is declined. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method of predicting and presenting a future question in an automated question answering system, the method performed by one or more hardware processors, comprising: receiving an input question in the automated question answering system; providing an answer to the question; searching a question database storing questions and a conditional probability of one question to be asked given that a previous question was asked, for all questions in the question database, the conditional probability determined based on a supervised learning algorithm, natural language processing that determines distance metric between the questions, and user data obtained from a plurality of data sources; predicting the future question based on the conditional probability stored in the question database given the input question as the previous question; suggesting the future question on a user interface; responsive to receiving an acceptance of the future question, updating the question database to strengthen the conditional probability associated with the future question occurring given the input question; providing an answer to the future question; repeating the searching, the predicting, the suggesting and the updating, with the future question as the input question until the future question is declined. 2. The method of claim 1, wherein the question database is structured as a directed weighted graph comprising nodes representing the questions and edges between the nodes representing links based on conditional probabilities between the questions. 3. The method of claim 1, wherein the question database is customized for a particular user in the particular user's current context, and the searching the question database comprises searching the customized question database. 4. The method of claim 1, wherein the question database is initially built using domain knowledge, and the supervised learning algorithm is trained with a training data set comprising the initially built question database, the trained supervised learning algorithm building the question database based on additional available data. 5. The method of claim 4, wherein the natural language processing that determines distance metric between the questions, and the user data obtained from a plurality of data sources extend the training data set and the supervised learning algorithm is automatically retrained based on the extended training data set. 6. The method of claim 1, wherein responsive to not finding the answer to the input question, searching an external network for the answer to the input question, adding the input question to the question database, and correlating the input question with the conditional probability based on at least one of the supervised learning algorithm, the natural language processing and the user data. 7. The method of claim 1, wherein the user data comprises social network data. 8. The method of claim 1, wherein the user data comprises user interaction history with the automated question answering system. 9. A computer readable storage medium storing a program of instructions executable by a machine to perform a method of predicting and presenting a future question in an automated question answering system, the method comprising: receiving an input question in the automated question answering system; providing an answer to the question; searching a question database storing questions and a conditional probability of one question to be asked given that a previous question was asked, for all questions in the question database, the conditional probability determined based on a supervised learning algorithm, natural language processing that determines distance metric between the questions, and user data obtained from a plurality of data sources; predicting the future question based on the conditional probability stored in the question database given the input question as the previous question; suggesting the future question on a user interface; responsive to receiving an acceptance of the future question, updating the question database to strengthen the conditional probability associated with the future question occurring given the input question; providing an answer to the future question; repeating the searching, the predicting, the suggesting and the updating, with the future question as the input question until the future question is declined. 10. The computer readable storage medium of claim 9, wherein the question database is structured as a directed weighted graph comprising nodes representing the questions and edges between the nodes representing links based on conditional probabilities between the questions. 11. The computer readable storage medium of claim 9, wherein the question database is customized for a particular user in the particular user's current context, and the searching the question database comprises searching the customized question database. 12. The computer readable storage medium of claim 9, wherein the question database is initially built using domain knowledge, and the supervised learning algorithm is trained with a training data set comprising the initially built question database, the trained supervised learning algorithm building the question database based on additional available data. 13. The computer readable storage medium of claim 12, wherein the natural language processing that determines distance metric between the questions, and the user data obtained from a plurality of data sources extend the training data set and the supervised learning algorithm is automatically retrained based on the extended training data set. 14. The computer readable storage medium of claim 9, wherein responsive to not finding the answer to the input question, searching an external network for the answer to the input question, adding the input question to the question database, and correlating the input question with the conditional probability based on at least one of the supervised learning algorithm, the natural language processing and the user data. 15. The computer readable storage medium of claim 14, wherein the external network comprises Internet. 16. The computer readable storage medium of claim 9, wherein the user data comprises social network data, data obtained from a wearable device and global positioning system location data. 17. The computer readable storage medium of claim 9, wherein the user data comprises user interaction history with the automated question answering system. 18. A system of predicting and presenting a future question in an automated question answering system, comprising: at least one hardware processor coupled to a communication network; a storage device coupled to the at least one hardware processor; the at least one hardware processor operable to receive an input question in the automated question answering system; the at least one hardware processor further operable to provide an answer to the question; the at least one hardware processor further operable to search a question database storing in the storage device, questions and a conditional probability of one question to be asked given that a previous question was asked, for all questions in the question database, the conditional probability determined based on a supervised learning algorithm, natural language processing that determines distance metric between the questions, and user data obtained from a plurality of data sources; the at least one hardware processor further operable to predict the future question based on the conditional probability stored in the question database given the input question as the previous question; the at least one hardware processor further operable to suggest the future question on a user interface; responsive to receiving an acceptance of the future question, the at least one hardware processor further operable to update the question database to strengthen the conditional probability associated with the future question occurring given the input question; the at least one hardware processor further operable to provide an answer to the future question; the at least one hardware processor further operable to repeat the searching, the predicting, the suggesting and the updating, with the future question as the input question until the future question is declined. 19. The system of claim 18, wherein the question database is structured as a directed weighted graph comprising nodes representing the questions and edges between the nodes representing links based on conditional probabilities between the questions. 20. The system of claim 18, wherein the question database is customized for a particular user in the particular user's current context, and the searching the question database comprises searching the customized question database. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A question database storing questions and a conditional probability of one question to be asked given that a previous question was asked is searched to predict a future question based on the conditional probability stored in the question database given an input question as the previous question. The future question is suggested to a user. Responsive to receiving an acceptance of the future question, the question database is updated to strengthen the conditional probability associated with the future question occurring given the input question. An answer to the future question is provided and searching, predicting, suggesting and updating may be repeated, with the future question as the input question, until the future question is declined. |
|
G06N7005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A question database storing questions and a conditional probability of one question to be asked given that a previous question was asked is searched to predict a future question based on the conditional probability stored in the question database given an input question as the previous question. The future question is suggested to a user. Responsive to receiving an acceptance of the future question, the question database is updated to strengthen the conditional probability associated with the future question occurring given the input question. An answer to the future question is provided and searching, predicting, suggesting and updating may be repeated, with the future question as the input question, until the future question is declined. |
|
A first indication from a user is received. The indication includes a task to be performed using at least one application programming interface. A machine learning model is determine. At least one application programming interface is determined using the machine learning model and the request. The at least one application programming interface is provided to the user. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for predicting application programming interface storytelling mapping, the method comprising the steps of: receiving, by one or more computer processors, a first indication from a user, wherein the indication includes a task to be added to an application programming interface storytelling mapping, wherein storytelling mapping is an organized layout of multiple application programming interface to perform the task; determining, by one or more computer processors, a machine learning model; determining, by one or more computer processors, at least one application programming interface using the machine learning model and the task; providing, by one or more computer processors, the at least one application programming interface to the user; receiving, by one or more computer processors, a second indication from the user, wherein the second indication includes a change to the at least one application programming interface; and responsive to receiving the second indication from the user, updating, by the one or more computer processors, the machine learning model using the change to the at least one application programming interface. 2. The method of claim 1, wherein the step of determining, by one or more computer processors, a machine learning model comprise: creating, by one or more computer processors, one or more machine learning models using one or more of the following: information about at least one preset existing application programming interface created to perform at least one task, information about at least one existing application programing interface used in an existing at least one application and an associated task performed by the existing at least one application, information about an order of the at least one application programing interface used in the existing at least one applications, information about links between the at least one application programming interface, and information about how data is passed between the at least one application programming interface; and determining, by one or more computer processors, a machine learning model of the at least one or more machine models based on one or more of the following: the user or the task. 3. The method of claim 1, wherein the at least one application programming interface is in a first order to perform the task. 4. The method of claim 1, wherein the first indication is analyzed using natural language processing to determine the task. 5. The method of claim 2, further comprising: receiving, by one or more computer processors, at least one update to one or more of the following: information about at least one preset existing application programming interface created to perform at least one task, information about at least one existing application programing interface used in an existing at least one application and an associated task performed by the existing at least one application, information about an order of the at least one application programing interface used in the existing at least one applications, information about links between the at least one application programming interface, and information about how data is passed between the at least one application programming interface; and updating, by one or more computer processors, the one or more machine learning models based on the at least one update. 6. A computer program product for predicting application programming interface storytelling mapping, the computer program product comprising: one or more computer readable storage media; and program instructions stored on the one or more computer readable storage media, the program instructions comprising: program instructions to receive a first indication from a user, wherein the indication includes a task to be added to an application programming interface storytelling mapping, wherein storytelling mapping is an organized layout of multiple application programming interface to perform the task; program instructions to determine a machine learning model; program instructions to determine at least one application programming interface using the machine learning model and the task; program instructions to provide the at least one application programing interface to the user; program instructions to receive a second indication from the user, wherein the second indication includes a change to the at least one application programming interface; and program instructions, responsive to receiving the second indication, to update the machine learning model using the change to the at least one application programming interface. 7. The computer program product of claim 6, wherein the program instructions to determine a machine learning model comprise: program instructions to create one or more machine learning models using one or more of the following: information about at least one preset existing application programming interface created to perform at least one task, information about at least one existing application programing interface used in an existing at least one application and an associated task performed by the existing at least one application, information about an order of the at least one application programing interface used in the existing at least one applications, information about links between the at least one application programming interface, and information about how data is passed between the at least one application programming interface; and program instructions to determine a machine learning model of the at least one or more machine models based on one or more of the following: the user or the task. 8. The computer program product of claim 6, wherein the at least one application programming interface is in a first order to perform the task. 9. The computer program product of claim 6, wherein the first indication is analyzed using natural language processing to determine the task. 10. The computer program product of claim 7, further comprising program instructions, stored on the one or more computer readable storage media, to: receive at least one update to one or more of the following: information about at least one preset existing application programming interface created to perform at least one task, information about at least one existing application programing interface used in an existing at least one application and an associated task performed by the existing at least one application, information about an order of the at least one application programing interface used in the existing at least one applications, information about links between the at least one application programming interface, and information about how data is passed between the at least one application programming interface; and update the one or more machine learning models based on the at least one update. 11. A computer system for predicting application programming interface storytelling mapping, the computer system comprising: one or more computer processors; one or more computer readable storage media; and program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more computer processors, the program instructions comprising: program instructions to receive a first indication from a user, wherein the indication includes a task to be added to an application programming interface storytelling mapping, wherein storytelling mapping is an organized layout of multiple application programming interface to perform the task; program instructions to determine a machine learning model; program instructions to determine at least one application programming interface using the machine learning model and the task; program instructions to provide the at least one application programing interface to the user; program instructions to receive a second indication from the user, wherein the second indication includes a change to the at least one application programming interface; and program instructions, responsive to receiving the second indication, to update the machine learning model using the change to the at least one application programming interface. 12. The computer system of claim 11, wherein the program instructions to determine a machine learning model comprise: program instructions to create one or more machine learning models using one or more of the following: information about at least one preset existing application programming interface created to perform at least one task, information about at least one existing application programing interface used in an existing at least one application and an associated task performed by the existing at least one application, information about an order of the at least one application programing interface used in the existing at least one applications, information about links between the at least one application programming interface, and information about how data is passed between the at least one application programming interface; and program instructions to determine a machine learning model of the at least one or more machine models based on one or more of the following: the user or the task. 13. The computer system of claim 11, wherein the at least one application programming interface is in a first order to perform the task. 14. The computer system of claim 12, further comprising program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more computer processors, to: receive at least one update to one or more of the following: information about at least one preset existing application programming interface created to perform at least one task, information about at least one existing application programing interface used in an existing at least one application and an associated task performed by the existing at least one application, information about an order of the at least one application programing interface used in the existing at least one applications, information about links between the at least one application programming interface, and information about how data is passed between the at least one application programming interface; and update the one or more machine learning models based on the at least one update. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A first indication from a user is received. The indication includes a task to be performed using at least one application programming interface. A machine learning model is determine. At least one application programming interface is determined using the machine learning model and the request. The at least one application programming interface is provided to the user. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A first indication from a user is received. The indication includes a task to be performed using at least one application programming interface. A machine learning model is determine. At least one application programming interface is determined using the machine learning model and the request. The at least one application programming interface is provided to the user. |
|
Systems, methods, and non-transitory computer readable media can obtain a conversation of a user in a chat application associated with a system, where the conversation includes one or more utterances by the user. An analysis of the one or more utterances by the user can be performed. A sentiment associated with the conversation can be determined based on a machine learning model, wherein the machine learning model is trained based on a plurality of features including demographic information associated with users. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implemented method comprising: obtaining, by a computing system, a conversation of a user in a chat application associated with a system, the conversation including one or more utterances by the user; performing, by the computing system, an analysis of the one or more utterances by the user; and determining, by the computing system, a sentiment associated with the conversation based on a machine learning model, wherein the machine learning model is trained based on a plurality of features including demographic information associated with users. 2. The computer-implemented method of claim 1, wherein the system is a social networking system, and the conversation is between the user and an agent associated with a page of an entity in the social networking system. 3. The computer-implemented method of claim 1, further comprising training the machine learning model based on the plurality of features, wherein the plurality of features further include one or more of: attributes associated with conversations, attributes associated with post history of users, or attributes associated with page history of users. 4. The computer-implemented method of claim 1, wherein the machine learning model provides one or more of: a sentiment score associated with the conversation or a sentiment label associated with the conversation. 5. The computer-implemented method of claim 4, wherein the sentiment label is indicative of a rating on a rating scale. 6. The computer-implemented method of claim 1, wherein the sentiment associated with the conversation is determined in or near real time. 7. The computer-implemented method of claim 1, wherein the performing the analysis of the one or more utterances by the user includes performing a textual analysis of an utterance by the user. 8. The computer-implemented method of claim 1, wherein the performing the analysis of the one or more utterances by the user includes determining a sentiment associated with one or more of: an emoticon, an emoji, or an indicator relating to text style. 9. The computer-implemented method of claim 1, wherein the sentiment associated with the conversation is determined based at least in part on the analysis of the one or more utterances by the user. 10. The computer-implemented method of claim 1, further comprising: determining an updated sentiment associated with the conversation based on the machine learning model at a time subsequent to a time at which the sentiment is determined; and detecting a change between the sentiment and the updated sentiment. 11. A system comprising: at least one hardware processor; and a memory storing instructions that, when executed by the at least one processor, cause the system to perform: obtaining a conversation of a user in a chat application associated with a system, the conversation including one or more utterances by the user; performing an analysis of the one or more utterances by the user; and determining a sentiment associated with the conversation based on a machine learning model, wherein the machine learning model is trained based on a plurality of features including demographic information associated with users. 12. The system of claim 11, wherein the instructions further cause the system to perform training the machine learning model based on the plurality of features, wherein the plurality of features further include one or more of: attributes associated with conversations, attributes associated with post history of users, or attributes associated with page history of users. 13. The system of claim 11, wherein the machine learning model provides one or more of: a sentiment score associated with the conversation or a sentiment label associated with the conversation. 14. The system of claim 11, wherein the sentiment associated with the conversation is determined in or near real time. 15. The system of claim 11, wherein the sentiment associated with the conversation is determined based at least in part on the analysis of the one or more utterances by the user. 16. A non-transitory computer readable medium including instructions that, when executed by at least one hardware processor of a computing system, cause the computing system to perform a method comprising: obtaining a conversation of a user in a chat application associated with a system, the conversation including one or more utterances by the user; performing an analysis of the one or more utterances by the user; and determining a sentiment associated with the conversation based on a machine learning model, wherein the machine learning model is trained based on a plurality of features including demographic information associated with users. 17. The non-transitory computer readable medium of claim 16, wherein the method further comprises training the machine learning model based on the plurality of features, wherein the plurality of features further include one or more of: attributes associated with conversations, attributes associated with post history of users, or attributes associated with page history of users. 18. The non-transitory computer readable medium of claim 16, wherein the machine learning model provides one or more of: a sentiment score associated with the conversation or a sentiment label associated with the conversation. 19. The non-transitory computer readable medium of claim 16, wherein the sentiment associated with the conversation is determined in or near real time. 20. The non-transitory computer readable medium of claim 16, wherein the sentiment associated with the conversation is determined based at least in part on the analysis of the one or more utterances by the user. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Systems, methods, and non-transitory computer readable media can obtain a conversation of a user in a chat application associated with a system, where the conversation includes one or more utterances by the user. An analysis of the one or more utterances by the user can be performed. A sentiment associated with the conversation can be determined based on a machine learning model, wherein the machine learning model is trained based on a plurality of features including demographic information associated with users. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Systems, methods, and non-transitory computer readable media can obtain a conversation of a user in a chat application associated with a system, where the conversation includes one or more utterances by the user. An analysis of the one or more utterances by the user can be performed. A sentiment associated with the conversation can be determined based on a machine learning model, wherein the machine learning model is trained based on a plurality of features including demographic information associated with users. |
|
One or more embodiments of a content naming system provide machine-learned name suggestions to a user for naming content items. Specifically, an online content management system can train a machine-learning model to identify a naming pattern from previously stored content items corresponding to a user account of the user. The online content management system uses the machine-learning model to determine a plurality of name suggestions for naming a content item associated with the user account. One or more embodiments provide graphical elements corresponding to the name suggestions within a graphical user interface. The user can select one or more graphical elements to add the corresponding name suggestion(s) to the name of the content item. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method comprising: training, based on characteristics of a plurality of previously named content items corresponding to a user account, a machine-learning model to identify a naming pattern associated with the plurality of previously named content items; determining, by at least one processor and using the machine-learning model, a plurality of name suggestions for a content item associated with the user account; providing, for display in a graphical user interface, a plurality of graphical elements corresponding to the plurality of name suggestions for naming the content item; identifying a selected graphical element from the plurality of graphical elements; and adding a name suggestion corresponding to the selected graphical element to the name of the content item. 2. The method as recited in claim 1, wherein determining the plurality of name suggestions for the content item comprises: determining one or more characteristics of the content item; and using the machine-learning model to determine the plurality of name suggestions based on the one or more characteristics of the content item. 3. The method as recited in claim 1, further comprising: preselecting one or more graphical elements from the plurality of graphical elements; and auto-populating the name of the content item with one or more corresponding name suggestions. 4. The method as recited in claim 1, further comprising enabling, for display in the graphical user interface, a modification of the name of the content item by adding, removing, or rearranging one or more graphical elements corresponding to one or more name suggestions. 5. The method as recited in claim 1, wherein the plurality of name suggestions comprise a plurality of name segments. 6. The method as recited in claim 1, further comprising: identifying a plurality of selected graphical elements from the plurality of graphical elements; and adding a plurality of name suggestions corresponding to the plurality of selected graphical elements to the name of the content item. 7. The method as recited in claim 6, wherein adding the plurality of name suggestions corresponding to the plurality of selected graphical elements to the name of the content item comprises adding the plurality of name suggestions in an order corresponding to an order in which the plurality of selected graphical elements are selected. 8. The method as recited in claim 1, wherein: determining the plurality of name suggestions for the content item associated with the user account comprises generating, using the machine-learning model, a plurality of scores for the plurality of name suggestions; and providing the plurality of graphical elements corresponding to the plurality of name suggestions for naming the content item comprises selecting, for display in the graphical user interface and based on the plurality of scores for the plurality of name suggestions, the plurality of graphical elements. 9. The method as recited in claim 8, wherein selecting the plurality of graphical elements comprises selecting a predetermined number of graphical elements. 10. The method as recited in claim 8, wherein selecting the plurality of graphical elements comprises selecting graphical elements corresponding to name suggestions having scores that meet a predetermined threshold. 11. The method as recited in claim 1, further comprising training the machine-learning model to identify a naming pattern associated with a plurality of user accounts associated with a related group of users. 12. The method as recited in claim 1, further comprising: detecting a text input comprising one or more manually entered characters; and determining the plurality of name suggestions based on the text input and the naming pattern. 13. A system comprising: at least one processor; and a non-transitory computer readable storage medium comprising instructions that, when executed by the at least one processor, cause the system to: identify characteristics of a plurality of previously named content items corresponding to a user account; train, based on characteristics of the plurality of previously named content items corresponding to the user account, a machine-learning model to identify a naming pattern associated with the plurality of previously named content items; determine, by the at least one processor and using the machine-learning model, a name segment for a new content item associated with the user account; provide, for display in a graphical user interface, a graphical element corresponding to the name segment for naming the new content item; and add, in response to a selection of the graphical element, the name segment to the name of the new content item. 14. The system as recited in claim 13, further comprising instructions that, when executed by the at least one processor, cause the system to determine the name segment for the new content item by: determining a characteristic of the new content item; and determining, using the identified naming pattern, the name segment according to the determined characteristic of the new content item. 15. The system as recited in claim 14, further comprising instructions that, when executed by the at least one processor, cause the system to determine the name segment according to the determined characteristic of the new content item by: comparing the determined characteristic of the new content item to one or more characteristics of the plurality of previously named content items; and selecting a name segment from a plurality of possible name segments based on a similarity of the characteristic of the new content item to the one or more characteristics of the plurality of previously named content items. 16. The system as recited in claim 13, further comprising instructions that, when executed by the at least one processor, cause the system to auto-populate, using the machine-learning model, the name of the new content item with one or more name segments based on one or more characteristics of the new content item. 17. The system as recited in claim 16, further comprising instructions that, when executed by the at least one processor, cause the system to: provide, for display in the graphical user interface, an option to remove the one or more name segments auto-populating the name of the new content item; and modify the name of the new content item in response to a selection of the option to remove the one or more name segments from the name of the new content item. 18. A non-transitory computer readable storage medium comprising instructions that, when executed by at least one processor, cause a computer system to: train, based on a plurality of content items previously stored in connection with a user account, a machine-learning model to identify a naming pattern associated with the plurality of content items; determine, by the at least one processor and using the machine-learning model, a plurality of name suggestions for a content item associated with the user account; provide, for display in a graphical user interface, a plurality of graphical elements corresponding to the plurality of name suggestions for naming the content item; identify a plurality of selected graphical elements from the plurality of graphical elements; and add a plurality of name suggestions corresponding to the plurality of selected graphical elements to the name of the content item. 19. The non-transitory computer readable storage medium of claim 18, further comprising instructions that, when executed by the at least one processor, cause the computer system to: determine the plurality of name suggestions for the content item by generating, using the machine-learning model, a plurality of scores for the plurality of name suggestions; and provide the plurality of graphical elements corresponding to the plurality of name suggestions for naming the content item by selecting, for display in the graphical user interface and based on the plurality of scores for the plurality of name suggestions, a plurality of graphical elements. 20. The non-transitory computer readable storage medium of claim 18, further comprising instructions that, when executed by the at least one processor, cause the computer system to train, based on a plurality of content items for a plurality of related user accounts, a machine-learning model to identify a naming pattern associated with the plurality of related user accounts. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: One or more embodiments of a content naming system provide machine-learned name suggestions to a user for naming content items. Specifically, an online content management system can train a machine-learning model to identify a naming pattern from previously stored content items corresponding to a user account of the user. The online content management system uses the machine-learning model to determine a plurality of name suggestions for naming a content item associated with the user account. One or more embodiments provide graphical elements corresponding to the name suggestions within a graphical user interface. The user can select one or more graphical elements to add the corresponding name suggestion(s) to the name of the content item. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: One or more embodiments of a content naming system provide machine-learned name suggestions to a user for naming content items. Specifically, an online content management system can train a machine-learning model to identify a naming pattern from previously stored content items corresponding to a user account of the user. The online content management system uses the machine-learning model to determine a plurality of name suggestions for naming a content item associated with the user account. One or more embodiments provide graphical elements corresponding to the name suggestions within a graphical user interface. The user can select one or more graphical elements to add the corresponding name suggestion(s) to the name of the content item. |
|
To simplify assisting a user in their day-to-day activities, a communication for performing an action may be sent to a user in the form of a query, where the query includes the most likely set of choices for the action arranged in a group of dichotomous (e.g., yes/no) or multiple choice answers. In this manner, a user may respond to the query by simply selecting one of the dichotomous or multiple choice answers. Historical logs of past actions, responses, queries, and so forth, may be used to predict future user actions or needs, and to formulate future queries for sending to the user. These techniques may be implemented, for example, through a remote coordination server or directly through a user's personal electronics device. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method comprising: receiving a communication at a personal electronic device; analyzing the communication to determine a set of choices in response to the communication, where at least one choice from the set of choices has an associated task predicted for a user of the personal electronic device; creating, by a processor of the personal electronic device, a query from the set of choices, the query including at least one dichotomous or multiple choice question arranged in a hierarchical structure to select among the set of choices, the hierarchical structure having a root and one or more leaves, where the root of the hierarchical structure corresponds to the at least one choice, and the one or more leaves of the hierarchical structure correspond to the associated task predicted for the user of the personal electronic device; presenting the at least one dichotomous or multiple choice question to the user of the personal electronic device in a user interface of the personal electronic device; receiving an input in response to the at least one dichotomous or multiple choice question through the user interface; based on the input, traversing the hierarchical structure, by the processor, to reduce the set of choices to a selected choice; based on the selected choice, where the selected choice is the at least one choice from the set of choices having the associated task, determining one or more actions to accomplish the associated task; obtaining information for performing the one or more actions; and performing the one or more actions using the personal electronic device. 2. The method of claim 1, wherein the root and the one or more leaves are on a same level of the hierarchical structure. 3. The method of claim 1, wherein the root is on a first level of the hierarchical structure and the one or more leaves are on one or more second levels of the hierarchical structure that are different than the first level. 4. The method of claim 1, wherein the analyzing the communication is based on a user data source associated with the user of the personal electronic device. 5. The method of claim 4, wherein the user data source includes a historical log of past actions taken by the user of the personal electronic device, and where the analyzing includes discerning one or more patterns from the historical log of past actions to determine the set of choices. 6. The method of claim 1, wherein the analyzing the communication is based on an external data source associated with the user of the personal electronic device for inferring patterns of activity, behavior, or intent from the user of the personal electronic device. 7. The method of claim 1, wherein the analyzing the communication is based on a data store of historical user data for groups of users. 8. The method of claim 1 wherein the query further comprises multiple dichotomous or multiple choice questions, wherein each dichotomous or multiple choice question narrows down a decision made by the user of the personal electronic device for performing the associated task. 9. The method of claim 1, wherein the personal electronic device is a smartphone, a watch, a headset, or a vehicle. 10. A computer program product comprising non-transitory computer executable code embodied in a non-transitory computer readable medium that, when executing on a personal electronics device, performs the steps of: receiving a communication at the personal electronic device; analyzing the communication to determine a set of choices in response to the communication, where at least one choice from the set of choices has an associated task predicted for a user of the personal electronic device; creating, by a processor of the personal electronic device, a query from the set of choices, the query including at least one dichotomous or multiple choice question arranged in a hierarchical structure to select among the set of choices, the hierarchical structure having a root and one or more leaves, where the root of the hierarchical structure corresponds to the at least one choice, and the one or more leaves of the hierarchical structure correspond to the associated task predicted for the user of the personal electronic device; presenting the at least one dichotomous or multiple choice question to the user of the personal electronic device in a user interface of the personal electronic device; receiving an input in response to the at least one dichotomous or multiple choice question through the user interface; based on the input, traversing the hierarchical structure, by the processor, to reduce the set of choices to a selected choice; based on the selected choice, where the selected choice is the at least one choice from the set of choices having the associated task, determining one or more actions to accomplish the associated task; obtaining information for performing the one or more actions; and performing the one or more actions using the personal electronic device. 11. The computer program product of claim 10, wherein the root and the one or more leaves are on a same level of the hierarchical structure. 12. The computer program product of claim 10, wherein the analyzing the communication is based on a user data source associated with the user of the personal electronic device. 13. The computer program product of claim 10, wherein the analyzing the communication is based on an external data source associated with the user of the personal electronic device for inferring patterns of activity, behavior, or intent from the user of the personal electronic device. 14. The computer program product of claim 10, wherein the analyzing the communication is based on a data store of historical user data for groups of users. 15. The computer program product of claim 10 wherein the query further comprises multiple dichotomous or multiple choice questions, wherein each dichotomous or multiple choice question narrows down a decision made by the user of the personal electronic device for performing the associated task. 16. The computer program product of claim 10, wherein the personal electronic device is a smartphone, a watch, a headset, or a vehicle. 17. A personal electronic device comprising: a processor; a memory; and a user interface, the processor coupled to the memory and to the user interface, the processor configured to: receive a communication, analyze the communication to determine a set of choices in response to the communication, where at least one choice from the set of choices has an associated task predicted for a user, create a query from the set of choices, the query including at least one dichotomous or multiple choice question arranged in a hierarchical structure to select among the set of choices, the hierarchical structure having a root and one or more leaves, where the root of the hierarchical structure corresponds to the at least one choice, and the one or more leaves of the hierarchical 4structure correspond to the associated task, present, on the user interface, the at least one dichotomous or multiple choice question to the user, receive, at the user interface, an input in response to the at least one dichotomous or multiple choice question, based on the input, traverse the hierarchical structure to reduce the set of choices to a selected choice; based on the selected choice, where the selected choice is the at least one choice from the set of choices having the associated task, determine one or more actions to accomplish the associated task; obtain information for performing the one or more actions; and perform the one or more actions using the personal electronic device. 18. A method comprising: receiving a communication at a personal electronic device; analyzing the communication to determine a set of choices in response to the communication, where at least one choice from the set of choices has an associated task predicted for a user of the personal electronic device; selecting, by a processor of the personal electronic device, the at least one choice based on at least one data source; determining, by the processor, one or more actions to accomplish the associated task; obtaining, by the processor, information for performing the one or more actions; and performing the one or more actions using the personal electronic device. 19. The method of claim 18, wherein the at least one data source includes current circumstances, activities, and known data about the user. 20. The method of claim 18, further comprising: creating, by the processor of the personal electronic device, a confirmation query from the selected choice, the confirmation query including at least one dichotomous question arranged in a hierarchical structure, the hierarchical structure having a root and one or more leaves, where the root of the hierarchical structure corresponds to the at least one choice, and the one or more leaves of the hierarchical structure correspond to the associated task; presenting the at least one dichotomous question to the user of the personal electronic device in a user interface of the personal electronic device; receiving an input in response to the at least one dichotomous question through the user interface; and based on the input, either confirming the selected choice; or selecting another choice based on the at least one data source. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: To simplify assisting a user in their day-to-day activities, a communication for performing an action may be sent to a user in the form of a query, where the query includes the most likely set of choices for the action arranged in a group of dichotomous (e.g., yes/no) or multiple choice answers. In this manner, a user may respond to the query by simply selecting one of the dichotomous or multiple choice answers. Historical logs of past actions, responses, queries, and so forth, may be used to predict future user actions or needs, and to formulate future queries for sending to the user. These techniques may be implemented, for example, through a remote coordination server or directly through a user's personal electronics device. |
|
G06N7005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: To simplify assisting a user in their day-to-day activities, a communication for performing an action may be sent to a user in the form of a query, where the query includes the most likely set of choices for the action arranged in a group of dichotomous (e.g., yes/no) or multiple choice answers. In this manner, a user may respond to the query by simply selecting one of the dichotomous or multiple choice answers. Historical logs of past actions, responses, queries, and so forth, may be used to predict future user actions or needs, and to formulate future queries for sending to the user. These techniques may be implemented, for example, through a remote coordination server or directly through a user's personal electronics device. |
|
A learn-by-example (LBE) system comprises, among other things, a first component which provides examples of data of interest (Supply Component/Example Data component); a second component capable of selecting and configuring a classification algorithm to classify the collected data (Configuration Component), and a third component capable of using the configured classification algorithm to classify new data from the sensors (Recognition Component). Together, these components detect sensory events of interest utilizing an LBE methodology, thereby enabling continuous sensory processing without the need for specialized sensor processing expertise and specialized domain-specific algorithm development. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A learning system for automatically detecting events of interest by processing data collected from one or more physical sensors in a user device, the system comprising: a first component that retrieves examples of events of interest from sensor data collected from at least one physical sensor; a second component that receives the examples of events of interest from the first component, and using a processor, classifies the examples into a plurality of categories to create a configured classification algorithm capable of categorizing subsequent events of interest; and a third component that executes the configured classification algorithm to compare newly available sensor data from the user device with the previously available examples of events of interest, and, upon the occurrence of an event of interest, determines an appropriate category of that particular event of interest detected in the newly available sensor data. 2. The system of claim 1, wherein the third component generates an output signal that performs a task in the user device. 3. The system of claim 1, wherein the first component, the second component and the third component are physically arranged in the user device itself. 4. The system of claim 1, wherein at least one of the first, second and third components is not physically arranged in the user device, but is communicatively coupled to the user device via wired or wireless connectivity. 5. The system of claim 1, wherein the collection of sensor data is affirmatively initiated by a user by a gesture-based, tactile, or audio command, or a combination thereof. 6. The system of claim 1, wherein the collection of sensor data is automatically initiated by an application that detects a behavioral pattern of a user before, during or after the occurrence of an event of interest. 7. The system of claim 6, wherein a circular buffer in a trace collector in the first component collects traces of events of interest involving an action by a user a part of the user's behavioral pattern. 8. The system of claim 7, wherein the collected traces are used for training the system. 9. The system of claim 8, wherein a feedback loop informs a person an estimated accuracy of automatic detection of events of interest. 10. The system of claim 1, wherein the examples of events of interest are retrieved from one or more of: a remote database, a local database, and, a circular buffer of a trace collector that temporarily collects incoming sensor data to detect potential events of interest. 11. The system of claim 1, wherein the configured classification algorithm created by the second component is based on neural networking techniques. 12. The system of claim 1, wherein a software application can selectively enable or disable distortion of data used as input for the classification algorithm. 13. The system of claim 12, wherein available forms of data distortion include one or more of amplitude distortion, frequency distortion, coordinate translation, mirror translation, velocity distortion, rotational distortion, variation of sensor data sampling rate, compression, and expansion. 14. The system of claim 1, wherein the third component is configured to adjust, automatically or via user feedback, the configured classification algorithm to generate a customized output. 15. The system of claim 14, wherein the adjustment of the configured classification algorithm includes changing parameters of the configured classification algorithm to ensure better match with an example event of interest. 16. The system of claim 14, wherein the customized output includes a confidence level for recognizing one or more events of interest. 17. The system of claim 14, wherein the customized output includes identification of a plurality of events of interest detected simultaneously, wherein each event of interest is classified into a corresponding appropriate category. 18. The system of claim 17, wherein the customized output further includes respective confidence levels for recognizing each of the plurality of events of interest, or a combined confidence level. 19. A computer-implemented method for automatically detecting events of interest by processing data collected from one or more physical sensors in a user device, the method comprising: retrieving examples of events of interest from sensor data collected from at least one physical sensor; receiving the retrieved examples of events of interest, and using a processor, classifying the examples into a plurality of categories to create a configured classification algorithm capable of categorizing subsequent events of interest; and executing the configured classification algorithm to compare newly available sensor data from the user device with the previously available examples of events of interest, and, upon the occurrence of an event of interest, determining an appropriate category of that particular event of interest detected in the newly available sensor data. 20. The method of claim 19, wherein the method further includes: generating an output signal that performs a task in the user device. 21. The method of claim 19, wherein the configured classification algorithm is based on neural networking techniques. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A learn-by-example (LBE) system comprises, among other things, a first component which provides examples of data of interest (Supply Component/Example Data component); a second component capable of selecting and configuring a classification algorithm to classify the collected data (Configuration Component), and a third component capable of using the configured classification algorithm to classify new data from the sensors (Recognition Component). Together, these components detect sensory events of interest utilizing an LBE methodology, thereby enabling continuous sensory processing without the need for specialized sensor processing expertise and specialized domain-specific algorithm development. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A learn-by-example (LBE) system comprises, among other things, a first component which provides examples of data of interest (Supply Component/Example Data component); a second component capable of selecting and configuring a classification algorithm to classify the collected data (Configuration Component), and a third component capable of using the configured classification algorithm to classify new data from the sensors (Recognition Component). Together, these components detect sensory events of interest utilizing an LBE methodology, thereby enabling continuous sensory processing without the need for specialized sensor processing expertise and specialized domain-specific algorithm development. |
|
Aspects herein describe new methods of determining optimal actions to achieve high-level goals with minimum total future cost. At least one high-level goal is inputted into a user device along with various observational data about the world, and a computational unit determines, through particle methods, an optimal course of action as well as emotions. The particle method comprises alternating backward and forward sweeps and tests for convergence to determine said optimal course of action. In one embodiment a user inputs a high-level goal into a cell phone which senses observational data. The cell phone communicates with a server that provides instructions. The server determines an optimal course of action via the particle method, and the cell phone then displays the instructions and emotions to the user. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method comprising: receiving, by a sensor, a current observation about a real or simulated world; maintaining, by a computational unit, an objective; representing the objective using an incremental cost of a plurality of potential actions; maintaining, by the computational unit, a current uncertainty about an unknown state of the world, wherein the current uncertainly is updated using the current observation; and determining, by the computational unit, one or more optimal actions to achieve the objective with an optimized expected total future cost, wherein said determining comprises performing both backward induction on the optimized expected total future cost and forward induction on the uncertainty about the unknown state of the world. 2. The method of claim 1, further comprising: determining a first number t, which represents a future time; determining a first vector x, which represents the unknown state of the world at time t; determining a second vector u, which represents a potential action at time t; determining a sequence of vectors i, called the information, which represents the preferred history of observations from the current time up to time t, leading to state x; determining a first function ƒ(t,x), called the local cost function; determining a first probability distribution p(t,x), called the unconditional uncertainty, which represents the uncertainty of the plurality of unknown states x of the world at time t, unconditional on future observations; wherein said induction comprises: restricting the set of future uncertainties conditional on observations that need to be considered to probability distributions of the form pi(t,x), which comprises the restriction of the unconditional uncertainty p(t,x) to states x having the same information i at time t; representing the expected total future cost at time t with information i as Epi[f], where the operator E[ ] represents the expectation of a function with respect to a probability distribution; determining the local cost function ƒ by backward induction; determining the unconditional uncertainty p and the information i by forward induction; and determining the optimal actions u by optimization for each value of the information i and time t. 3. The method of claim 2, wherein the steps of representing the objective and the stochastic dynamics of the world include: determining second number dt, which represents the time step; determining a third vector y, which represents the observation at time t; determining a first stochastic function A(t,x,u), called the dynamics, whose values provide the random distribution of state after time step dt when starting from state x and taking the potential action u at time t; determining a second stochastic function B(t,x,u), called the measurement, whose values provide the random distribution of observations after time step dt when starting from state x and taking the potential action u at time t; and determining a second function C(t,y,u), which represents the incremental cost over time step dt of the potential action u, based in part on the observation y, at time t, wherein backward induction determines fat time t from time t+dt by: f(t,x)=C(t,y,u)+Ex′˜A(t,x,u)[f(t+dt,x′)] where ˜ denotes the distribution of a stochastic variable, wherein forward induction determines p at time t+dt from time t by: p(t+dt,x′)˜{x′˜A(t,x,u)|x˜p(t,x)} and determines i at time t+dt from time t as a preferred value of i appended with the observation y, and wherein the optimization of the potential actions determines u at time t by: u=argminEpi[f(t,x)|u]. 4. The method of claim 3, further comprising determining an approximate solution of the backward and forward induction equations by means of an unstructured sample, comprising: determining a sample of particles such that each particle in the sample possesses: an instance of time t, an instance of the unknown state x, an instance of the observation y, an instance of the potential action u, an instance of the value of the local cost function ƒ, an instance of the information i, an instance of a weight representing the probability assigned by the unconditional uncertainty p, and an instance of a uniform weight q representing the relative number of states of the world represented by this sampled particle; repeating, until convergence is achieved, the following unordered steps: initializing the instance of the value f of the local cost function of each particle to a known terminal cost, performing a backward sweep in time of the instance of the value f of the local cost function on the sampled particles, determining a candidate for the optimal potential action u at the current time, initializing the instance of the weight p of each particle to approximate the known current uncertainty about the unknown state of the world, initializing the instance of the information i of each particle to the known current observation y, performing a forward sweep in time of the instance of the weight p and the instance of the information i on the sampled particles, determining whether the instance of the weight p and the instance of the information i have converged for all particles at a terminal time, or whether the candidate for the optimal action u has converged at the initial time; and identifying one or more optimal actions as candidate optimal actions in the final iteration of this sequence of steps when convergence has been achieved. 5. The method of claim 4, wherein performing the backward sweep comprises repetition of the following unordered steps until the current time is reached: propagating the particles backward in time; performing particle interactions; and optionally resampling the particles, and wherein the forward sweep comprises repetition of the following unordered steps until the terminal time is reached: propagating the particles forward in time; performing particle interactions; and optionally resampling the particles. 6. The method of claim 5 wherein the computations involved in each step in respect of each particle is carried out by means of a parallel computational sub-unit, and the particle interactions are effected by means of communication between the parallel computational sub-units. 7. The method of claim 5 wherein the sampling or resampling particles further includes: determining a desired un-normalized particle sampling density r(x); and employing importance sampling according to the desired un-normalized sampling density r(x). 8. The method of claim 7 wherein the step of employing importance sampling according to the un-normalized density r(x) comprises applying, one or more times, the following sequence of steps to all particles in the old sample to produce the new sample: starting with an old particle with state x; randomly drawing a symmetric jump from state x to another state x′; creating, with a probability of min(1,r(x′)/r(x)) a new particle with state x′; creating, independently and with probability max(0,1−r(x′)/r(x)), a new particle with state x. 9. The method of claim 3, wherein the computation of the variance of the total future cost under the assumption of optimal actions is obtained as: Epi[f2]−(Epi[f])2, wherein the local square cost function ƒ2 (t,x) is determined by backward induction according to the equation: f2(t,x)=C(t,y,u)2+2C(t,y,u)Ex′˜A(t,x,u)[f(t+dt,x′)]+Ex′˜A(t,x,u)[f2(t+dt,x′)]. 10. The method of claim 2, wherein the steps of representing the objective and the stochastic dynamics of the world include: determining a third vector y, which represents the observation at time t; determining a second function c(t,y,u), which represents the incremental cost of the potential action u, based in part on the observation y, at time t; determining a third function a(t,x,u), called the expected dynamics, which represents the average rate of change of the state x when the potential action u is taken at time y; determining a fourth function s(t,x), called the noise, which represents the standard deviation of the rate of change of the state x at time t; and determining a fifth function m(t,x), called the measurement, which represents the measurement process that yields the observation y in state x at time t. 11. The method of claim 10, wherein the local cost function ƒ is determined by backward induction according to the backward time differential equation: - f . = c + f x a + 1 2 ( f ) xx s 2 ; the unconditional uncertainty p is determined by forward induction according to the forward time differential equation: p t = - p x a - p tr ( a x + a u u x ) + 1 2 p xx s 2 ; and wherein for each i, a representation of a potential action, u, is an optimal action and t is determined by the equation: u=cu−1(−Epi[aufu]). 12. The method of claim 11, further including determining an approximate solution of the backward and forward differential equations, comprising: determining a sample of particles such that each particle in the sample possesses: an instance of time t, an instance of the unknown state x, an instance of the observation y, an instance of the potential action u, an instance of the value of the local cost function ƒ, an instance of the information i, an instance of a weight representing the probability assigned by the unconditional uncertainty distribution p, an instance of a uniform weight q representing an approximation of a uniform probability distribution on unknown states at time t; repeating the following sequence of steps until convergence is achieved: initializing the instance of the value f of the local cost function of each particle to a known terminal cost, performing a backward sweep in time of the instance of the value f of the local cost function on the sampled particles, determining a candidate for the optimal potential action u at the current time, initializing the instance of the weight p of each particle to approximate the known current uncertainty about the unknown state of the world, initializing the instance of the information i of each particle to the known current observation y, performing a forward sweep in time of the instance of the weight p and the instance of the information i on the sampled particles, determining whether the instance of the weight p and the instance of the information i have converged for all particles at a terminal time; and identifying one or more optimal actions as candidate optimal actions in the final iteration of this sequence of steps when convergence has been achieved. 13. The method of claim 12 wherein performing the backward sweep comprises repetition of the following steps until the current time is reached: propagating the particles backward in time; performing particle interactions; and optionally resampling the particles, and wherein the forward sweep comprises repetition of the following steps until the terminal time is reached: propagating the particles forward in time; performing particle interactions; and optionally resampling the particles. 14. A method of simulating an emotional response, comprising: receiving, by a sensor, a current observation about a real or simulated world; maintaining, by a computational unit, an objective; representing the objective using an incremental cost of a plurality of potential actions; maintaining, by the computational unit, a current uncertainty about an unknown state of the world, wherein the current uncertainly is updated using the current observation; determining, by the computational unit, one or more optimal actions to achieve the objective with an optimized expected total future cost; determining, via the computational unit, a statistic of the computation determining the optimal actions, corresponding to a simulated emotion; and communicating a representation of the emotion to the world. 15. The method of claim 14, wherein the statistic of the computation is the optimized expected total future cost, of which low values correspond to a happy emotion and high values correspond to a sad emotion. 16. The method of claim 14, wherein the statistic of the computation is the variance of the total future cost under an assumption of optimal actions, of which low values correspond to a calm emotion and high values correspond to a tense emotion, wherein a combination of tense and sad emotions is interpreted as a fearful emotion, and wherein a combination of tense and happy emotions is interpreted as an excited emotion. 17. The method of claim 14, wherein the statistic of the computation is the covariance of the total future costs of two competing objectives under an assumption of optimal actions, the first objective representing the self, and the second objective representing another actor in the world, of which a negative value in combination with a sad emotion is interpreted as an angry emotion, and a negative value in combination with a happy emotion is interpreted as a gloating emotion, and a positive value in combination with a happy emotion is interpreted as a loving emotion. 18. A system comprising: a sensor configured to receive input defining a current observation about a real or simulated world; a computational unit configured to: maintain an objective, represent the objective using an incremental cost of a plurality of potential actions, maintain a current uncertainty about an unknown state of the world, wherein the current uncertainly is updated using the current observation, and determine one or more optimal actions to achieve the objective, wherein said determining comprises performing both backward induction on the minimum expected total future cost and forward induction on the uncertainty about the unknown state of the world. 19. The system of claim 18, further comprising an actuator configured to effect the one or more optimal actions based on instructions from the computational unit. 20. The method of claim 1, further comprising effecting, via an actuator, the one or more optimal actions. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: Aspects herein describe new methods of determining optimal actions to achieve high-level goals with minimum total future cost. At least one high-level goal is inputted into a user device along with various observational data about the world, and a computational unit determines, through particle methods, an optimal course of action as well as emotions. The particle method comprises alternating backward and forward sweeps and tests for convergence to determine said optimal course of action. In one embodiment a user inputs a high-level goal into a cell phone which senses observational data. The cell phone communicates with a server that provides instructions. The server determines an optimal course of action via the particle method, and the cell phone then displays the instructions and emotions to the user. |
|
G06N5042 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Aspects herein describe new methods of determining optimal actions to achieve high-level goals with minimum total future cost. At least one high-level goal is inputted into a user device along with various observational data about the world, and a computational unit determines, through particle methods, an optimal course of action as well as emotions. The particle method comprises alternating backward and forward sweeps and tests for convergence to determine said optimal course of action. In one embodiment a user inputs a high-level goal into a cell phone which senses observational data. The cell phone communicates with a server that provides instructions. The server determines an optimal course of action via the particle method, and the cell phone then displays the instructions and emotions to the user. |
|
A data processing system is disclosed for machine learning. The system comprises a sampling module (13) and a computational module (15) interconnected by a data communications link (17). The computational module is configured to store a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units. The sampling module is configured to receive the parameter vector from the first processing module and to sample from the probability distribution defined by the parameter vector to produce state vectors for the network. The computational module is further configured to receive the state vectors from the second processing module and to apply an algorithm to produce new data. The sampling and computational modules are configured to operate independently from one another. | Please help me write a proper abstract based on the patent claims. CLAIM: 1-48. (canceled) 49. A method comprising: storing, by a first processing module, a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units; and receiving, by the first processing module, the state vectors from a second processing module and applying an algorithm to produce new data, wherein the state vectors produced for the network by the second processing module are sampled from a probability distribution defined by the parameter vector produced by the first processing module; and wherein the sampling is performed independently from the producing the new data. 50. A method according to claim 49, comprising sampling and producing the new data at least partially in parallel. 51. A method according to claim 49, comprising sampling and/or producing the new data until a predetermined condition is reached, at which time data is exchanged from one processing module to the other processing module. 52. A method according to claim 51, wherein applying the algorithm comprises applying a learning algorithm to produce an updated parameter vector that is subsequently sent to the second processing module for re-sampling when the predetermined condition is reached. 53. A method according to claim 52, comprising sending the updated parameter vector to the second processing module for re-sampling when a predetermined plural number of iterations of the learning algorithm have been performed. 54. A method according to claim 49, wherein the algorithm takes as input the state vectors and the parameter vector to generate output data representing a probability distribution. 55. A method according to claim 49, comprising buffering the samples received prior to applying the algorithm. 56. A method according to claim 49, comprising using at least two samplers to generate the state vectors and, prior to applying the algorithm, receiving the sampled output from each and removing duplicates. 57. A method according to claim 49, wherein the second processing module is implemented as a quantum annealing machine. 58. A system comprising: a first processing module configured to store a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units; and wherein the first processing module is further configured to receive the state vectors from a second processing module and to apply an algorithm to produce new data, wherein the state vectors produced for the network by the second processing module are sampled from a probability distribution defined by the parameter vector produced by the first processing module. wherein the processing modules are configured to sample and produce the new data independently from one another. 59. A system according to claim 58, wherein the processing modules are configured to sample and produce the new data at least partially in parallel. 60. A system according to claim 58, wherein the processing modules are configured to sample and/or produce the new data until a predetermined condition is reached, at which time data is exchanged between the two processing modules. 61. A system according to claim 60, wherein the first processing module is configured to apply a learning algorithm to produce an updated parameter vector and subsequently to send it to the second processing module for re-sampling when the predetermined condition is reached. 62. A system according to claim 61, wherein the first processing module is configured to send the updated parameter vector to the second processing module for re-sampling when a predetermined plural number of iterations of the learning algorithm have been performed. 63. A system according to claim 58, wherein the first processing module is configured to take as input the state vectors and the parameter vector, and to generate output data representing a probability distribution. 64. A system according to claim 58, wherein the first processing module is configured to buffer the samples received prior to applying the algorithm. 65. A system according to claim 58, wherein the second processing module comprises at least two samplers configured to generate the state vectors and, prior to applying the algorithm, to receive the sampled output from each and to remove duplicates. 66. A system according to claim 58, wherein the first and second processing modules are implemented on different hardware modules having their own microprocessor or microcontroller. 67. A system according to claim 58, wherein the second processing module is implemented as a quantum annealing machine. 68. A non-transitory computer-readable storage medium having stored thereon computer-readable code, which, when executed by computing apparatus, causes the computing apparatus to perform a method comprising: storing, by a first processing module, a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units; and receiving, by the first processing module, the state vectors from a second processing module and applying an algorithm to produce new data, wherein the state vectors produced by the second processing module are sampled from a probability distribution defined by the parameter vector produced by the first processing module; and wherein the sampling is performed independently from the producing the new data. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A data processing system is disclosed for machine learning. The system comprises a sampling module (13) and a computational module (15) interconnected by a data communications link (17). The computational module is configured to store a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units. The sampling module is configured to receive the parameter vector from the first processing module and to sample from the probability distribution defined by the parameter vector to produce state vectors for the network. The computational module is further configured to receive the state vectors from the second processing module and to apply an algorithm to produce new data. The sampling and computational modules are configured to operate independently from one another. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A data processing system is disclosed for machine learning. The system comprises a sampling module (13) and a computational module (15) interconnected by a data communications link (17). The computational module is configured to store a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units. The sampling module is configured to receive the parameter vector from the first processing module and to sample from the probability distribution defined by the parameter vector to produce state vectors for the network. The computational module is further configured to receive the state vectors from the second processing module and to apply an algorithm to produce new data. The sampling and computational modules are configured to operate independently from one another. |
|
An apparatus and method are provided for spatial processing of concepts. Included is a non-transitory memory comprising a distance matrix and instructions, where the distance matrix includes values representing dissimilarities among a plurality of concepts stored in a knowledgebase. Further included is one or more processors in communication with the memory. The one or more processors execute the instructions to derive an inner product matrix, based on the distance matrix. Further, a spectral decomposition of the inner product matrix is performed. Based on the spectral decomposition of the inner product matrix, a plurality of concept vectors are generated. Information associated with the plurality of concept vectors is then output for spatial processing. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A processing device, comprising: a non-transitory memory comprising a distance matrix and instructions, with the distance matrix including values representing dissimilarities among a plurality of concepts stored in a knowledgebase; one or more processors in communication with the memory, wherein the one or more processors execute the instructions to: derive an inner product matrix based on the distance matrix; perform a spectral decomposition of the inner product matrix; generate a plurality of concept vectors based on the spectral decomposition of the inner product matrix; and output information associated with the plurality of concept vectors for spatial processing. 2. The processing device of claim 1, wherein the spectral decomposition of the inner product matrix is configured such that the plurality of concept vectors are generated with a dimension that is equal to a rank of the inner product matrix. 3. The processing device of claim 1, wherein the spectral decomposition of the inner product matrix is configured such that the plurality of concept vectors are generated with a dimension that is less than a rank of the inner product matrix. 4. The processing device of claim 1, wherein the one or more processors further execute the instructions to: receive user input including a dimension, wherein the plurality of concept vectors are generated with the dimension based on the spectral decomposition of the inner product matrix. 5. The processing device of claim 1, wherein the spatial processing includes displaying the information associated with the plurality of concept vectors. 6. The processing device of claim 1, wherein the spatial processing includes displaying the information associated with the plurality of concept vectors in Euclidean space. 7. The processing device of claim 1, wherein a concept of the plurality of concepts includes at least one word, a phrase, or a plurality of words of a sentence. 8. The processing device of claim 7, wherein the one or more processors further execute the instructions to assemble the plurality of concept vectors into a concept matrix representative of the sentence. 9. The processing device of claim I, wherein the spatial processing includes at least one of deep learning, text analysis, or clustering. 10. The processing device of claim 1, wherein the information associated with the plurality of concept vectors includes one or both of the plurality of concept vectors or information derived from the plurality of concept vectors. 11. A method, comprising: a processing device deriving an inner product matrix based on a distance matrix, with the distance matrix including values representing dissimilarities among a plurality of concepts stored in a knowledgebase; the processing device performing a spectral decomposition of the inner product matrix; the processing device generating a plurality of concept vectors based on the spectral decomposition of the inner product matrix; and the processing device outputting information associated with the plurality of concept vectors for spatial processing. 12. The method of claim 13, further comprising the processing device receiving user input including a dimension, wherein the plurality of concept vectors are generated with the dimension based on the spectral decomposition of the inner product matrix. 13. The method of claim 13, wherein the spectral decomposition of the inner product matrix is configured such that the plurality of concept vectors are generated with a dimension that is equal to a rank of the inner product matrix. 14. The method of claim 13, wherein the spectral decomposition of the inner product matrix is configured such that the plurality of concept vectors are generated with a dimension that is less than a rank of the inner product matrix. 15. The method of claim 13, wherein the spatial processing includes displaying the information associated with the plurality of concept vectors. 16. The method of claim 13, wherein the spatial processing includes displaying the information associated with the plurality of concept vectors in Euclidean space. 17. The method of claim 13, wherein a concept of the plurality of concepts includes at least one word, a phrase, or a plurality of words of a sentence. 18. The method of claim 17, further comprising assembling the plurality of concept vectors into a concept matrix representative of the sentence. 19. The method of claim 13, wherein the spatial processing includes at least one of deep learning, text analysis, or clustering. 20. The method of claim 13, wherein the information associated with the plurality of concept vectors includes one or both of the plurality of concept vectors or information derived from the plurality of concept vectors. 21. A non-transitory computer-readable media storing computer instructions, that when executed by one or more processors, cause the one or more processors to perform the steps of: deriving an inner product matrix based on a distance matrix, with the distance matrix including values representing dissimilarities among a plurality of concepts stored in a knowledgebase; performing a spectral decomposition of the inner product matrix; generating a plurality of concept vectors based on the spectral decomposition of the inner product matrix; and outputting information associated with the plurality of concept vectors for spatial processing. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: An apparatus and method are provided for spatial processing of concepts. Included is a non-transitory memory comprising a distance matrix and instructions, where the distance matrix includes values representing dissimilarities among a plurality of concepts stored in a knowledgebase. Further included is one or more processors in communication with the memory. The one or more processors execute the instructions to derive an inner product matrix, based on the distance matrix. Further, a spectral decomposition of the inner product matrix is performed. Based on the spectral decomposition of the inner product matrix, a plurality of concept vectors are generated. Information associated with the plurality of concept vectors is then output for spatial processing. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: An apparatus and method are provided for spatial processing of concepts. Included is a non-transitory memory comprising a distance matrix and instructions, where the distance matrix includes values representing dissimilarities among a plurality of concepts stored in a knowledgebase. Further included is one or more processors in communication with the memory. The one or more processors execute the instructions to derive an inner product matrix, based on the distance matrix. Further, a spectral decomposition of the inner product matrix is performed. Based on the spectral decomposition of the inner product matrix, a plurality of concept vectors are generated. Information associated with the plurality of concept vectors is then output for spatial processing. |
|
A user may be exposed to multiple token instances representing stimuli that may influence the affective state of the user. Described herein are embodiments of systems, method, and computer programs for estimating affective response to a token instance of interest, selected from among the token instances. The token instance of interest is selected based on attention levels computed utilizing a model for predicting interest in token instances. In one example embodiment, the token instance of interest is a token instance for which attention level of the user is higher than at least one other token instance. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A system configured to utilize predicted attention levels to estimate affective response to a token instance of interest, comprising: a memory configured to store: a total affective response of a user to token instances to which the user was exposed and representations of first and second attention levels in respective first and second token instances from among the token instances; wherein the first and second attention levels are computed utilizing a model for predicting interest in token instances; wherein the first and second token instances have overlapping instantiation periods; and wherein the first attention level is greater than the second attention level; a processor configured to select, based on the representations of the first and second attention levels, the first token instance as the token instance of interest; and the processor is further configured to estimate the affective response to the token instance of interest from the total affective response. 2. The system of claim 1, wherein the total affective response is based, at least in part, on a value of a user measurement channel of the user. 3. The system of claim 1, wherein the first and second token instances are visual token instances, and the model for predicting interest in token instances comprises an algorithm for prediction of interest in images. 4. The system of claim 1, wherein the first and second attention levels are computed utilizing a machine learning-based predictor that utilizes the model to predict interest levels in token instances; wherein the predictor is trained on data comprising attention levels of the user to token instances to which the user was previously exposed. 5. The system of claim 1, wherein the model comprises predictions of interest of other users in the first and second token instances, and the first and second attention levels are computed by combining the predictions using a machine learning-based collaborative filtering approach. 6. The system of claim 1, wherein the total affective response is expressed as an emotional response. 7. The system of claim 1, wherein the total affective response is expressed as a value of a user measurement channel of the user. 8. The system of claim 1, wherein the total affective response is proportional to the difference between affective responses of the user before and after the user was exposed to the token instances. 9. The system of claim 1, wherein the memory is further configured to store a baseline value for affective response of the user, and the processor is further configured to calculate the total affective response with respect to the baseline value; whereby the baseline value represents a usual affective response of the user, as computed from multiple values acquired over a period of at least a few hours. 10. The system of claim 1, wherein more than 50% of the total affective response is attributed to the affective response to the token instance of interest. 11. A method for utilizing predicted attention levels to estimate affective response to a token instance of interest, comprising: receiving a total affective response of a user to token instances to which the user was exposed; wherein the token instances comprise first and second token instances having overlapping instantiation periods; receiving representations of first and second attention levels in the first and second token instances, respectively; wherein the first and second attention levels are computed utilizing a model for predicting interest in token instances; and wherein the first attention level is greater than the second attention level; selecting, based on the representations of the first and second attention levels, the first token instance as the token instance of interest; and estimating the affective response to the token instance of interest from the total affective response. 12. The method of claim 11, wherein the first and second token instances are visual token instances, and the model for predicting interest in token instances comprises an algorithm for prediction of interest in images. 13. The method of claim 11, further comprising computing the first and second attention levels utilizing a machine learning-based predictor that utilizes the model to predict interest levels in token instances; and further comprising training the predictor on data comprising attention levels of the user to token instances to which the user was previously exposed. 14. The method of claim 11, wherein the model comprises predictions of interest of other users in the first and second token instances, and further comprising computing the first and second attention levels by combining the predictions using a machine learning-based collaborative filtering approach. 15. The method of claim 11, wherein the total affective response is proportional to the difference between affective responses of the user before and after the user was exposed to the token instances. 16. The method of claim 11, further comprising receiving a baseline value for affective response of the user, and calculating the total affective response with respect to the baseline value; whereby the baseline value represents a usual affective response of the user, as computed from multiple values acquired over a period of at least a few hours. 17. A non-transitory computer-readable medium for use in a computer to utilize predicted attention levels to estimate affective response to a token instance of interest; the computer comprises a processor, and the non-transitory computer-readable medium comprising: program code for receiving a total affective response of a user to token instances to which the user was exposed; wherein the token instances comprise first and second token instances having overlapping instantiation periods; program code for receiving representations of first and second attention levels in the first and second token instances, respectively; wherein the first and second attention levels are computed utilizing a model for predicting interest in token instances; and wherein the first attention level is greater than the second attention level; program code for selecting, based on the representations of the first and second attention levels, the first token instance as the token instance of interest; and program code for estimating the affective response to the token instance of interest from the total affective response. 18. The non-transitory computer-readable medium of claim 17, wherein the first and second token instances are visual token instances, and further comprising program for computing the first and second attention levels with an algorithm that utilizes the model for prediction of interest in images. 19. The non-transitory computer-readable medium of claim 17, further comprising program code for computing the first and second attention levels utilizing a machine learning-based predictor that utilizes the model to predict interest levels in token instances; and further comprising program code for training the predictor on data comprising attention levels of the user to token instances to which the user was previously exposed. 20. The non-transitory computer-readable medium of claim 17, wherein the model comprises predictions of interest of other users in the first and second token instances, and further comprising program code for computing the first and second attention levels by combining the predictions using a machine learning-based collaborative filtering approach. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: A user may be exposed to multiple token instances representing stimuli that may influence the affective state of the user. Described herein are embodiments of systems, method, and computer programs for estimating affective response to a token instance of interest, selected from among the token instances. The token instance of interest is selected based on attention levels computed utilizing a model for predicting interest in token instances. In one example embodiment, the token instance of interest is a token instance for which attention level of the user is higher than at least one other token instance. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A user may be exposed to multiple token instances representing stimuli that may influence the affective state of the user. Described herein are embodiments of systems, method, and computer programs for estimating affective response to a token instance of interest, selected from among the token instances. The token instance of interest is selected based on attention levels computed utilizing a model for predicting interest in token instances. In one example embodiment, the token instance of interest is a token instance for which attention level of the user is higher than at least one other token instance. |
Subsets and Splits