CommunityRuleChatBot/chatbot/routes.py

import pandas as pd
from flask_cors import CORS
from flask import (
    Blueprint, request, jsonify
)

bp = Blueprint('routes', __name__,)
df = pd.read_csv('rule.csv')
user_input = "Improve the living and working conditions of everyone everywhere Fair and humane laws and practices"
compare_column1 = df['summary']
compare_column2 = df['modules']
return_column = df['name']
@bp.route('/request', methods=('GET', 'POST'))
def handle_request():

    #user_input = request.args.get('question')
    #closest_value, ruleId = find_closest_match(user_input, df['summary'], df['modules'], df['name'], df['ruleID'])
    

    #finalValue = "https://communityrule.info/create/?r=" + str(ruleId)
    #response_message = f"{closest_value}"

    #response_data = {
    #    'community': response_message,
    #    'link': finalValue
    #}

    #return jsonify(response_data)
    
    user_input = request.args.get('question')
    top_matches = find_closest_match(user_input, df['summary'], df['modules'], df['name'], df['ruleID'])
    
    recommendations = []
    for similarity, name, ruleId in top_matches:
        finalValue = f"https://communityrule.info/create/?r={ruleId}"
        recommendations.append({
            'community': name,
            'link': finalValue,
        })

    print(jsonify(recommendations))
    return jsonify(recommendations)


    #finalValue = "https://communityrule.info/create/?r=" + str(ruleId)
    #response_message = f"{closest_value}"

    #response_data = {
    #    'community': response_message,
    #    'link': finalValue
    #}

    #return jsonify(response_data)

def find_closest_match(user_input, compare_column1_values, compare_column2_values, return_column_values, rule_id_values):
    print(user_input)
    max_similarity_curr = -1
    closest_value = None
    rule_id = None
    user_tokens = user_input.split()
    results = []
    # Compare user input with each value in the compare columns
    for compare_value1, compare_value2, return_value, ruleId in zip(compare_column1_values, compare_column2_values, return_column_values, rule_id_values):
        # Convert compare_value2 to string if it's a float
        if isinstance(compare_value1, float):
            compare_value1 = str(compare_value1)
        if isinstance(compare_value2, float):
            compare_value2 = str(compare_value2)

        # Split compare_value2 into tokens
        compare_tokens1 = compare_value1.split()
        compare_tokens2 = compare_value2.split()

        # Calculate similarity between user input and compare_value1 and compare_value2
        similarity1 = similarity_score(user_tokens, compare_tokens1)
        similarity2 = similarity_score(user_tokens, compare_tokens2)

        # Take the maximum similarity between similarity1 and similarity2
        max_similarity_curr = max(similarity1, similarity2)
        print(similarity1, similarity2)
        # Update closest value if current similarity is greater
        #if max_similarity_curr > max_similarity:
            #max_similarity = max_similarity_curr
            #closest_value = return_value
            #rule_id = ruleId
        results.append((max_similarity_curr, return_value, ruleId))
        # print(results)
    results.sort(reverse=True, key=lambda x: x[0])
    return results[:3]

def similarity_score(set1, set2):
    # Calculate the intersection of the two sets
    intersection = len(set(set1).intersection(set(set2)))
    
    # Calculate the union of the two sets
    union = len(set(set1).union(set(set2)))
    
    # Calculate the Jaccard similarity coefficient
    if union == 0:
        return 0
    else:
        return intersection / union