Apppublic

thevikramrajput/ai-feedback-analyzer

sourceHugging Faceupdated 6mo agoView on Hugging Face

0likes

App README

AI Product Feedback Analyzer

A production-style AI system that converts raw user reviews into actionable product insights using NLP and pre-trained models.

🔗 Live Demo: Hugging Face Spaces

🎯 Problem Statement

Product teams cannot manually read thousands of user reviews. They need an automated system to:

—Understand overall user sentiment
—Discover the main complaint categories
—Prioritize issues based on frequency and impact
—Get AI-generated summaries of what users are saying

🚀 Solution

An end-to-end AI pipeline that:

1.Loads Google Play Store reviews from CSV
2.Preprocesses text (cleaning, filtering English reviews)
3.Analyzes sentiment based on star ratings
4.Clusters negative reviews to discover complaint patterns
5.Generates AI summaries using HuggingFace Inference API (BART-large-CNN)
6.Visualizes everything in a product analytics dashboard

✨ Features

Feature	Description
Sentiment Analysis	Classify reviews as positive, neutral, or negative
Topic Clustering	Keyword-based clustering for complaint categories
AI Summarization	HuggingFace Inference API (BART-large-CNN) generates summaries
Category-wise Reviews	View ALL reviews in each complaint category
Interactive Dashboard	Professional Streamlit UI with charts and filters
Filter Support	Filter by sentiment, app, and view all reviews

🤖 AI-Powered Summary

Click "Generate Detailed AI Summary" to get:

—Overall Summary - AI-generated summary of all reviews
—What Users Are Complaining About - Summary of negative feedback
—What Users Love - Summary of positive reviews
—Category-wise Summaries - Login issues, bugs, feature requests, messaging
—Actual User Quotes - Real complaints and praises
—Action Items - Recommendations based on user feedback

🛠️ Tech Stack

Technology	Purpose
Python 3.9+	Core language
Streamlit	Dashboard UI
FastAPI	Backend API (local)
Pandas, NumPy	Data processing
scikit-learn	KMeans clustering
HF Inference API	BART-large-CNN for AI summarization
Plotly	Interactive charts

📁 Project Structure

ai-feedback-analyzer/
├── app.py               # Main Streamlit app (HF Spaces)
├── requirements.txt     # Dependencies
├── data/
│   └── Training_Data.csv
├── backend/             # FastAPI (local dev)
│   ├── app.py
│   ├── data_loader.py
│   ├── preprocess.py
│   ├── sentiment.py
│   ├── clustering.py
│   └── insights.py
└── frontend/
    └── app.py           # Streamlit (local dev)

⚡ Quick Start

Hugging Face Spaces (Recommended)

Visit: https://huggingface.co/spaces/thevikramrajput/ai-feedback-analyzer

Local Development

bash

# Install dependencies
pip install -r requirements.txt

# Run Streamlit app
streamlit run app.py

With Backend API (Optional)

bash

# Terminal 1: Start API
cd backend
uvicorn app:app --reload --port 8000

# Terminal 2: Start Dashboard
cd frontend
streamlit run app.py

📊 Dashboard Features

—KPI Cards - Total reviews, positive/negative %, health score
—Sentiment Pie Chart - Visual sentiment distribution
—Top Issues Bar Chart - Most frequent complaint categories
—AI Summary Section - Detailed insights from reviews
—Reviews by Category - View ALL reviews in each category
—All Reviews Table - Complete filterable review list

📝 CSV Format

Your CSV should have these columns:

—content - Review text (required)
—score - Star rating 1-5 (required)
—userLang - Language code (optional)
—app_id - App identifier (optional)

🔮 Future Improvements

—[ ] Multilingual support
—[ ] Time-series trend analysis
—[ ] Competitor comparison
—[ ] Export to PDF/Excel
—[ ] Custom LLM integration