app.py

import streamlit as st
import pandas as pd
from db import insert_data_if_empty, get_mongo_client
from chatbot import chatbot_response

# Ensure historical data is inserted into MongoDB if not already present.
insert_data_if_empty()

# Connect to MongoDB (optional: for additional visualizations)
collection = get_mongo_client()

st.subheader("💬 Chatbot with Sentiment Analysis & Category Extraction")
# Updated hint to include examples for basic questions and entry queries.
user_prompt = st.text_area(
    "Ask me something (e.g., 'Provide analysis for data entry 1 in the dataset' or 'What is the dataset summary?'):"
)

if st.button("Get Response"):
    ai_response, sentiment_label, sentiment_confidence, topic_label, topic_confidence = chatbot_response(user_prompt)
    if ai_response:
        st.write("### Response:")
        st.write(ai_response)
        st.write("### Sentiment Analysis:")
        # Convert sentiment confidence to percentage format (e.g., 70% confidence)
        st.write(f"**Sentiment Detected:** {sentiment_label} ({sentiment_confidence * 100:.0f}% confidence)")
        st.write("### Category Extraction:")
        st.write(f"**Category Detected:** {topic_label} ({topic_confidence * 100:.0f}% confidence)")
    else:
        st.warning("⚠️ Please enter a question or text for analysis.")