Sentiment Analysis Web Application

A modern full-stack Sentiment Analysis application built using React, Flask, and Hugging Face Transformers. The application allows users to analyze individual sentences or batches of text, visualize sentiment distribution through interactive charts, and generate AI-enhanced summary reports.

---

Overview

This project combines a responsive React frontend with a Flask backend powered by a Transformer-based sentiment analysis model. Users can classify text in real time, analyze large collections of feedback, and explore insights through interactive visualizations.

---

Features

Single Sentence Analysis

• Analyze individual sentences instantly.
• Predicts:
  • Positive
  • Negative
• Displays confidence score.

---

Batch Sentiment Analysis

Upload or paste multiple sentences to:

• Analyze each sentence individually.
• Group results into categories.
• Count positive and negative sentiments.
• View category-wise statistics.

---

AI Generated Summary Report

Generate a comprehensive report including:

• Overall sentiment summary
• Positive vs Negative percentages
• Category-wise insights
• Automatically generated narrative

---

Interactive Dashboard

Includes:

• Responsive interface
• Category filters
• Interactive Bar Charts
• Interactive Pie Charts
• Sentence-wise result table
• Loading animations
• Theme selection
• Axis/Grid visibility toggle

---

Production Ready

• Flask WSGI support
• CORS enabled
• Easily deployable on:
  • Render
  • Railway
  • Heroku
  • Docker
  • VPS

---

Tech Stack

Layer	Technology
Frontend	React 19
Build Tool	Vite 6
Styling	Tailwind CSS
Animation	Framer Motion
Charts	Chart.js
Backend	Flask
ML Framework	Hugging Face Transformers
Deep Learning	PyTorch
API	REST
Deployment	Gunicorn (WSGI)

---

Project Structure

Sentiment-Analysis/
│
├── README.md
├── package.json
├── vite.config.js
├── tailwind.config.js
├── eslint.config.js
├── index.html
├── wsgi.py
│
├── public/
│   └── vite.svg
│
├── src/
│   ├── main.jsx
│   ├── App.jsx
│   ├── App.css
│   ├── index.css
│   │
│   ├── Components/
│   │   ├── InputForm.jsx
│   │   ├── BarChart.jsx
│   │   ├── PieChart.jsx
│   │   ├── Report.jsx
│   │   ├── FilterDropdown.jsx
│   │   └── SentenceItem.jsx
│   │
│   ├── backend/
│   │   ├── app.py
│   │   ├── model.py
│   │   └── requirements.txt
│   │
│   └── assets/
│       └── react.svg
│
└── .gitignore

---

API Endpoints

1. Classify a Sentence

POST /classify

Request

{
    "text": "The product is amazing!"
}

Response

{
    "label": "Positive",
    "confidence": 0.998
}

---

2. Batch Analysis

POST /analyze

Request

{
    "sentences": [
        "Great service",
        "Very disappointing",
        "Loved it"
    ]
}

Response

{
    "results":[...],
    "summary":{
        "Positive":2,
        "Negative":1
    }
}

---

3. Generate Report

POST /generate-full-report

Returns an AI-generated narrative report summarizing:

• Overall sentiment
• Positive percentage
• Negative percentage
• Category-wise observations
• Final conclusion

---

Installation

Clone Repository

git clone https://github.com/AdityaPanda0506/Sentimental-Analysis.git

cd Sentimental-Analysis

---

Frontend Setup

Install dependencies

npm install

Start development server

npm run dev

Frontend runs at

http://localhost:5173

---

Backend Setup

Create virtual environment (optional)

Windows

python -m venv venv

venv\Scripts\activate

Linux / macOS

python3 -m venv venv

source venv/bin/activate

Install Python dependencies

pip install -r src/backend/requirements.txt

Run Flask server

python src/backend/app.py

Backend runs at

http://localhost:5000

---

Production Build

Build React

npm run build

Run using Gunicorn

gunicorn wsgi:app

---

Workflow

User
   │
   ▼
React Frontend
   │
   ▼
Flask REST API
   │
   ▼
Transformer Model
(Hugging Face)
   │
   ▼
Prediction
   │
   ▼
JSON Response
   │
   ▼
Charts + Reports

---

How It Works

Step 1

The user enters a sentence or multiple feedback entries.

↓

Step 2

The React frontend sends a request to the Flask API.

↓

Step 3

The backend loads a pretrained Hugging Face Transformer model.

↓

Step 4

The model predicts:

• Sentiment
• Confidence Score

↓

Step 5

Results are returned as JSON.

↓

Step 6

React renders:

• Bar Chart
• Pie Chart
• Sentence Table
• AI Summary Report

---

Security

• Flask-CORS enabled for local development.
• Configure allowed origins before production deployment.
• Store API keys using environment variables.
• Never commit secrets to the repository.

---

Future Improvements

• User authentication
• PDF report export
• CSV upload support
• Dark mode
• Model selection
• Multi-language sentiment analysis
• Emotion detection
• Aspect-based sentiment analysis
• Historical analytics dashboard

---

Deployment

This application can be deployed using:

• Docker
• Render
• Railway
• Heroku
• AWS EC2
• DigitalOcean
• Azure App Service

---

Contributing

1. Fork the repository.
2. Create a new branch.

git checkout -b feature/new-feature

3. Commit your changes.

git commit -m "Add new feature"

4. Push your branch.

git push origin feature/new-feature

5. Open a Pull Request.

---

License

This project is released for educational and research purposes.

Feel free to modify, extend, and use it with proper attribution.

---

Author

Aditya Panda

Built with

• React
• Flask
• Hugging Face Transformers
• PyTorch
• Chart.js
• Tailwind CSS