Utils file updated

utils.py

@@ -1,3 +1,4 @@
+# Importing libraries
 from bs4 import BeautifulSoup
 import requests
 from langchain.schema import HumanMessage
@@ -7,11 +8,14 @@ from dotenv import load_dotenv
 import os
 from transformers import pipeline
 
+# Load environment variables
 load_dotenv()
 GROQ_API_KEY = os.getenv('GROQ_API_KEY')
 
-llm=ChatGroq(api_key=GROQ_API_KEY, model="llama-3.1-8b-instant")
+# Initialize the LLM model
+llm = ChatGroq(api_key=GROQ_API_KEY, model="llama-3.1-8b-instant")
 
+# Function to extract news titles and summaries from Economic Times
 def extract_titles_and_summaries(company_name, num_articles=10):
     url = f"https://economictimes.indiatimes.com/topic/{company_name}/news"
     try:
@@ -48,10 +52,11 @@ def extract_titles_and_summaries(company_name, num_articles=10):
         print(f"An error occurred: {e}")
         return []
 
+# Function to perform sentiment analysis on extracted news articles
 def perform_sentiment_analysis(news_data):
     from transformers import pipeline
     articles = news_data.get("Articles", [])
-    pipe = pipeline("text-classification", model="tabularisai/multilingual-sentiment-analysis",device=1)
+    pipe = pipeline("text-classification", model="tabularisai/multilingual-sentiment-analysis", device=1)
     sentiment_counts = {"Positive": 0, "Negative": 0, "Neutral": 0}
 
     for article in articles:
@@ -77,13 +82,15 @@ def perform_sentiment_analysis(news_data):
 
     return news_data, sentiment_counts
 
+# Function to extract topics from articles using Hugging Face model
 def extract_topics_with_hf(news_data):
     structured_data = {
         "Company": news_data.get("Company", "Unknown"),
         "Articles": []
     }
-    topic_pipe = pipeline("text-classification", model="valurank/distilroberta-topic-classification",device=1)
+    topic_pipe = pipeline("text-classification", model="valurank/distilroberta-topic-classification", device=1)
     articles = news_data.get("Articles", [])
+
     for article in articles:
         content = f"{article['Title']} {article['Summary']}"
         topics_result = topic_pipe(content, top_k=3)
@@ -98,10 +105,12 @@ def extract_topics_with_hf(news_data):
         })
     return structured_data
 
+# Function to generate a final sentiment summary using LLM
 def generate_final_sentiment(news_data, sentiment_counts):
     company_name = news_data["Company"]
     total_articles = sum(sentiment_counts.values())
     combined_summaries = " ".join([article["Summary"] for article in news_data["Articles"]])
+
     prompt = f"""
     Based on the analysis of {total_articles} articles about the company "{company_name}":
     - Positive articles: {sentiment_counts['Positive']}
@@ -109,22 +118,26 @@ def generate_final_sentiment(news_data, sentiment_counts):
     - Neutral articles: {sentiment_counts['Neutral']}
     The following are the summarized key points from the articles: "{combined_summaries}".
     Provide a single, concise summary that integrates the overall sentiment analysis and key news highlights while maintaining a natural flow. Explain its implications for the company's reputation, stock potential, and public perception.
-    Respond **ONLY** with a well-structured very concised and very short paragraph in plain text, focus on overall sentiment.
+    Respond **ONLY** with a well-structured very concise and short paragraph in plain text, focusing on overall sentiment.
     """
-    response = llm.invoke([HumanMessage(content=prompt)],max_tokens=200)
+
+    response = llm.invoke([HumanMessage(content=prompt)], max_tokens=200)
     final_sentiment = response if response else "Sentiment analysis summary not available."
-    return final_sentiment.content   # it's a string
+    return final_sentiment.content  # returns a string
 
+# Function to extract JSON response from LLM output
 def extract_json(response):
     try:
         return json.loads(response)
     except json.JSONDecodeError:
         return {}
-    
+
+# Function to compare articles based on common topics and sentiment variations
 def compare_articles(news_data, sentiment_counts):
     articles = news_data.get("Articles", [])
     all_topics = [set(article["Topics"]) for article in articles]
     common_topics = set.intersection(*all_topics) if all_topics else set()
+
     topics_prompt = f"""
     Analyze the following article topics and identify **only three** key themes that are common across multiple articles, 
     even if they are phrased differently. The topics from each article are:
@@ -133,10 +146,12 @@ def compare_articles(news_data, sentiment_counts):
     Respond **ONLY** with a JSON format:
     {{"CommonTopics": ["topic1", "topic2", "topic3"]}}
     """
+    
     response = llm.invoke([HumanMessage(content=topics_prompt)]).content
     contextual_common_topics = extract_json(response).get("CommonTopics", list(common_topics))[:3]  # Limit to 3 topics
 
     total_articles = sum(sentiment_counts.values())
+
     comparison_prompt = f"""
     Provide a high-level summary comparing {total_articles} news articles about "{news_data['Company']}":
     - Sentiment distribution: {sentiment_counts}
@@ -155,9 +170,12 @@ def compare_articles(news_data, sentiment_counts):
         ]
     }}
     """
+
     response = llm.invoke([HumanMessage(content=comparison_prompt)]).content
     coverage_differences = extract_json(response).get("Coverage Differences", [])
+
     final_sentiment = generate_final_sentiment(news_data, sentiment_counts)
+
     return {
         "Company": news_data["Company"],
         "Articles": articles,
@@ -173,4 +191,4 @@ def compare_articles(news_data, sentiment_counts):
             }
         },
         "Final Sentiment Analysis": final_sentiment
-    }
+    }