Upload 2 files

main.py

@@ -0,0 +1,205 @@
+# -*- coding: utf-8 -*-
+"""Main.ipynb
+
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/github/payal15604/ONDC-Test/blob/main/Main.ipynb
+"""
+
+# Commented out IPython magic to ensure Python compatibility.
+!pip install git+https://github.com/PrithivirajDamodaran/ZSIC.git
+!pip install transformers -U
+!pip install pyDecision
+!git clone https://github.com/payal15604/ONDC-Test
+# %cd ONDC-Test
+
+from zero_shot_text import zero_shot_text as ztext
+from zero_shot_image import zero_shot_image as zimg
+
+from text_summarizer import text_summarizer as t_sum
+from similarity_scoring import calculate_similarity as c_sim
+
+import numpy as np
+
+"""JSON FILE WILL COME"""
+
+import json
+import requests
+
+# Define the URL of your API endpoint
+url = "http://localhost:4500/api/datasender"
+
+def parse_json(json_data):
+    try:
+        # Convert boolean literals to uppercase
+        json_data = json_data.replace("true", "True").replace("false", "False")
+        # Parse JSON data
+        parsed_json = json.loads(json_data)
+        return parsed_json
+    except json.JSONDecodeError:
+        print("Error: Invalid JSON data")
+        return None
+
+try:
+    # Send a POST request to the API endpoint
+    response = requests.post(url)
+
+    # Check if the request was successful (status code 200)
+    if response.status_code == 200:
+        # Extract the JSON data from the response
+        received_data = response.text
+
+        # Parse the JSON data
+        parsed_data = parse_json(received_data)
+
+        if parsed_data:
+            # Process each received item
+            for item in parsed_data.get("Received data", []):
+                # Accessing fields from the item
+                name = item.get("name", "")
+                short_disc = item.get("short_desc", "")
+                long_disc = item.get("long_desc", "")
+                image = item.get("images", [])
+                symbol = item.get("symbol", "")
+                
+                # Example: Print the fields of each item
+                print("Name:", name)
+                print("Short Description:", short_disc)
+                print("Long Description:", long_disc)
+                print("Images:", image)
+                print("Symbol:", symbol)
+
+                # TODO: Process the fields in your ML model
+
+        else:
+            print("Error: Failed to parse JSON data")
+
+    else:
+        # Handle the case where the request was not successful
+        print("Error: Failed to fetch data from the API. Status code:", response.status_code)
+
+except Exception as e:
+    # Handle any exceptions that occur during the request
+    print("Error:", e)
+labels = ['coffee','tea','shampoo','face serum','bread','honey','soap','biscuit','milk','chocolate','juice']
+
+result = ztext(name, labels)
+
+def zero_shot_text_formatted(text, labels):
+  result = ztext(text, labels)
+
+  temp_text_sequence = result['labels']
+  temp_text_scores = result['scores']
+  formatted_output = f"{temp_text_sequence[0]} = {temp_text_scores[0]}"
+
+  print(formatted_output)
+  return formatted_output
+
+def zero_shot_image_formatted(img, labels):
+  result = zimg(img, labels)
+
+  result_score = result['scores'][0]
+  result_label = result['labels'][0]
+  img_res = f"{result_label} = {result_score}"
+
+  print(img_res)
+  return img_res
+
+# temp_name_score = zero_shot_text_formatted(name, labels)
+# temp_sdisc_score = zero_shot_text_formatted(short_disc, labels)
+# temp_ldisc_score = zero_shot_text_formatted(long_disc, labels)
+
+# name_sdisc_score = c_sim(temp_name_score, temp_sdisc_score, model="en_core_web_sm")
+# name_ldisc_score = c_sim(temp_name_score, temp_ldisc_score, model="en_core_web_sm")
+# name_sldisc_score = c_sim(temp_sdisc_score, temp_ldisc_score, model="en_core_web_sm")
+
+# print(name_sdisc_score)
+# print(name_ldisc_score)
+# print(name_sldisc_score)
+
+# temp_summary = t_sum(long_disc)
+# print(temp_summary)
+
+# print(temp_name_score)
+# print(temp_sdisc_score)
+# print(temp_ldisc_score)
+# print(temp_summary)
+
+"""## **SCORING**"""
+
+def name_disc_score(name, short_disc, long_disc, labels):
+  n_compute = zero_shot_text_formatted(name, labels)
+  sd_compute = zero_shot_text_formatted(short_disc, labels)
+  ld_compute = zero_shot_text_formatted(long_disc, labels)
+
+  n_sd_score = c_sim(n_compute, sd_compute, model="en_core_web_sm")
+  n_ld_score = c_sim(n_compute, ld_compute, model="en_core_web_sm")
+  sd_ld_score = c_sim(sd_compute, ld_compute, model="en_core_web_sm")
+
+  return n_sd_score, n_ld_score, sd_ld_score
+
+def name_symbol_score(name, symbol, labels):
+  n_compute = zero_shot_text_formatted(name, labels)
+  s_compute = zero_shot_image_formatted(symbol, labels)
+
+  n_s_score = c_sim(n_compute, s_compute, model="en_core_web_sm")
+
+  return n_s_score
+
+def name_image_score(name, image_list, labels):
+    n_compute = zero_shot_text_formatted(name, labels)
+    n_i_scores = []
+
+    for i in range(len(image_list)):
+        i_compute = zero_shot_image_formatted(image_list[i], labels)
+        n_i_score = c_sim(n_compute, i_compute)
+        n_i_scores.append(n_i_score)
+
+    return sum(n_i_scores)/len(image_list)  #average
+
+N_Sd_score, N_Ld_score, Sd_Ld_score = name_disc_score(name, short_disc, long_disc, labels)
+
+print((N_Sd_score + N_Ld_score + Sd_Ld_score)/3)
+
+N_S_score = name_symbol_score(name, symbol, labels)
+
+N_S_score = name_image_score(name, image, labels)
+print(N_S_score)
+
+# Required Libraries
+import numpy as np
+
+from pyDecision.algorithm import topsis_method
+
+
+
+# TOPSIS
+def topsis(name, long_disc, short_disc, image, symbol,labels):
+
+  # Weights
+  #[0.3,0.25, 0.2, 0.1, 0.15]
+  weights = [0.2,0.2,0.2,0.2,0.2]  #assigned manually
+  N_i_score = name_image_score(name, image, labels)
+  N_Sd_score, N_Ld_score, Sd_Ld_score = name_disc_score(name, short_disc, long_disc, labels)
+  N_S_score = name_symbol_score(name, symbol, labels)
+  # Load Criterion Type: 'max' or 'min'
+  criterion_type = ['max', 'max', 'max', 'max','max']
+
+  # Dataset
+  dataset = np.array([
+                  [N_S_score, N_i_score, N_Sd_score, N_Ld_score, Sd_Ld_score],
+                  [0.87, 0.9, 0.47, 0.46, 0.5],
+                  [1,0.67,0.57,0.56,0.8]#demo data
+                  ])
+
+  # Call TOPSIS
+  relative_closeness = topsis_method(dataset, weights, criterion_type, graph = False, verbose = True)
+  return relative_closeness
+
+relative_closeness=topsis(name, long_disc, short_disc, image, symbol,labels)
+print(relative_closeness)
+
+## Call TOPSIS
+#relative_closeness = topsis_method(dataset, weights, criterion_type, graph = False, verbose = True)
+

requirements.txt

@@ -0,0 +1,13 @@
+tensorflow
+transformers -U
+pipeline
+AutoTokenizer
+AutoModelForSequenceClassification
+T5ForConditionalGeneration
+spacy
+NLTK
+git+https://github.com/PrithivirajDamodaran/ZSIC.git
+ZSIC
+pyDecision
+numpy
+