Datasets:

Modalities:
Text
Formats:
json
Languages:
English
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
jen commited on
Commit
3c972a7
·
verified ·
1 Parent(s): 4e20a3a

Update the dataset card to provide a better description of the dataset.

Browse files
Files changed (1) hide show
  1. README.md +20 -7
README.md CHANGED
@@ -4,24 +4,37 @@ configs:
4
  - config_name: david
5
  data_files:
6
  - split: train
7
- path: "david/train.jsonl"
8
  - split: test
9
- path: "david/test.jsonl"
10
  - config_name: isabel
11
  data_files:
12
  - split: train
13
- path: "isabel/train.jsonl"
14
  - split: test
15
- path: "isabel/test.jsonl"
16
  - config_name: marcus
17
  data_files:
18
  - split: train
19
- path: "marcus/train.jsonl"
20
  - split: test
21
- path: "marcus/test.jsonl"
 
 
 
 
 
 
 
22
  ---
23
 
24
  # The Panza Emails dataset
25
 
26
- This dataset contains of three collections of emails of three authentic users, with personal information (names, places, etc.) replaced by other ones for donor privacy.
27
  Except for these changes, the language of the emails is genuine. The intention of this dataset is to allow researchers to study strategies for text personalization.
 
 
 
 
 
 
 
4
  - config_name: david
5
  data_files:
6
  - split: train
7
+ path: david/train.jsonl
8
  - split: test
9
+ path: david/test.jsonl
10
  - config_name: isabel
11
  data_files:
12
  - split: train
13
+ path: isabel/train.jsonl
14
  - split: test
15
+ path: isabel/test.jsonl
16
  - config_name: marcus
17
  data_files:
18
  - split: train
19
+ path: marcus/train.jsonl
20
  - split: test
21
+ path: marcus/test.jsonl
22
+ task_categories:
23
+ - text2text-generation
24
+ language:
25
+ - en
26
+ pretty_name: Panza email
27
+ size_categories:
28
+ - n<1K
29
  ---
30
 
31
  # The Panza Emails dataset
32
 
33
+ This dataset contains collections of emails of three authentic users (david, isabel, and marcus), with personal information (names, places, etc.) replaced by other ones for donor privacy.
34
  Except for these changes, the language of the emails is genuine. The intention of this dataset is to allow researchers to study strategies for text personalization.
35
+ The data was donated explicitly for this purpose. This dataset is ethically collected and fully licensed for research use. The dataset
36
+ is entirely in English, and all authors are highly proficient in the language; all messages are authored entirely by the donor, without any
37
+ collaborative writing or LLM use.
38
+
39
+ This data was used in the paper [**Panza: Design and Analysis of a Fully-Local Personalized Text Writing Assistant**](https://arxiv.org/abs/2407.10994) to fine-tune open-source models to impersonate a specific author.
40
+ Code to do so is available on [GitHub](https://github.com/IST-DASLab/PanzaMail).