Skip to content

README.old.md

Latest commit

 

History

History
326 lines (254 loc) · 8.07 KB

README.old.md

File metadata and controls

326 lines (254 loc) · 8.07 KB

FinSentiment Pro - AI-Powered Financial Sentiment Analysis

A premium, production-ready Financial Sentiment Analysis web application with real-time multi-dimensional sentiment tracking for stock market intelligence.

🚀 Features

  • Multi-Model AI: FinBERT sentiment analysis + emotion detection
  • Voice Input: Speak stock tickers using Web Speech API
  • Real-Time Data: Live news from News API and stock prices from Yahoo Finance
  • Interactive Dashboards:
    • Overall Sentiment Gauge
    • Timeline Charts (Sentiment + Price correlation)
    • Multi-dimensional Radar Charts
    • Source Comparison
    • Calendar Heatmap
    • Correlation Scatter Plots
    • AI-Generated Insights
  • Dark/Light Theme: Smooth theme transitions with localStorage persistence
  • Responsive Design: Works perfectly on mobile, tablet, and desktop
  • Accessibility: WCAG 2.1 AA compliant
  • Performance Optimized: Lazy loading, code splitting, memoization

📋 Prerequisites

  • Node.js (v18 or higher)
  • Python (v3.8 or higher)
  • pip (Python package manager)

🛠️ Installation

1. Install Frontend Dependencies

npm install

2. Install Backend Dependencies

cd backend
pip install -r requirements.txt

The backend uses:

  • FinBERT for financial sentiment analysis
  • yfinance for stock price data
  • News API for real-time news articles
  • Transformers for emotion detection

🔑 API Keys Setup

To use this application, you'll need to obtain free API keys from:

  1. News API - Get your free key at https://newsapi.org/
  2. Alpha Vantage - Get your free key at https://www.alphavantage.co/support/#api-key
  3. Hugging Face - Get your token at https://huggingface.co/settings/tokens
  4. Indian Stock API - Get your key at https://indianapi.in/

Setup Instructions

  1. Copy the example environment file:

    cd backend
cp .env.example .env

  2. Edit the .env file and add your API keys:

    NEWS_API_KEY=your_newsapi_key_here
ALPHA_VANTAGE_KEY=your_alphavantage_key_here
HF_API_KEY=your_huggingface_key_here
INDIAN_STOCK_API_KEY=your_indian_stock_api_key_here

Note: The .env file is gitignored and will not be committed to the repository for security.

🚀 Running the Application

Option 1: Manual Start

Terminal 1 - Start Backend:

cd backend
python app.py

The backend will start on http://localhost:5000

Terminal 2 - Start Frontend:

npm run dev

The frontend will start on http://localhost:3000

Option 2: Quick Start (Single Command)

# Windows PowerShell
.\start.ps1

Or use the start script from package.json:

npm start

📱 Usage

  1. Enter a Stock Ticker: Type a ticker symbol (e.g., AAPL, TSLA, GOOGL) or use voice input
  2. Select Time Period: Choose 24 Hours, 7 Days, or 30 Days
  3. Click Analyze: The app will fetch news and perform sentiment analysis
  4. Explore Results:
    • View overall sentiment score
    • Analyze sentiment trends over time
    • Compare different news sources
    • Read AI-generated insights
    • Export reports

⌨️ Keyboard Shortcuts

  • / - Focus search input
  • D - Toggle dark/light mode
  • V - Activate voice input
  • Esc - Close modals

🎨 Tech Stack

Frontend

  • React 18 with TypeScript
  • Tailwind CSS for styling
  • Framer Motion for animations
  • Recharts for charts and graphs
  • Lucide React for icons
  • Zustand for state management
  • Axios for API calls

Backend

  • Flask for REST API
  • FinBERT (ProsusAI/finbert) for financial sentiment
  • Transformers for emotion detection
  • yfinance for stock data
  • News API for news articles
  • NumPy for data processing

📊 API Endpoints

POST /api/analyze

Analyze sentiment for a stock ticker.

Request:

{
  "ticker": "AAPL",
  "days": 7
}

Response:

{
  "ticker": "AAPL",
  "companyName": "Apple Inc.",
  "overallScore": 67.5,
  "confidence": 89,
  "totalArticles": 247,
  "timeline": [...],
  "dimensions": {...},
  "sourceBreakdown": [...],
  "headlines": [...],
  "insights": {...}
}

GET /api/health

Health check endpoint.

🏗️ Project Structure

nlp/
├── backend/
│   ├── app.py              # Flask API server
│   └── requirements.txt    # Python dependencies
├── src/
│   ├── components/
│   │   ├── Header.tsx
│   │   ├── Hero.tsx
│   │   ├── Dashboard.tsx
│   │   ├── Footer.tsx
│   │   └── dashboard/
│   │       ├── SearchSection.tsx
│   │       ├── OverallSentimentGauge.tsx
│   │       ├── TimelineChart.tsx
│   │       ├── RadarChart.tsx
│   │       ├── SourceComparison.tsx
│   │       ├── HeadlinesFeed.tsx
│   │       ├── CalendarHeatmap.tsx
│   │       ├── AIInsights.tsx
│   │       ├── ExportPanel.tsx
│   │       └── CorrelationScatter.tsx
│   ├── store/
│   │   └── useStore.ts     # Zustand state management
│   ├── types/
│   │   └── index.ts        # TypeScript types
│   ├── App.tsx
│   ├── main.tsx
│   └── index.css
├── package.json
├── tsconfig.json
├── vite.config.ts
├── tailwind.config.js
└── README.md

🎯 Features Breakdown

Sentiment Analysis

  • Real-time sentiment scoring (-100 to +100)
  • Multi-dimensional analysis (6 metrics)
  • Confidence levels for each prediction
  • Emotion detection (joy, fear, anger, surprise, etc.)

Data Visualization

  • Interactive timeline charts
  • Radar charts for multi-dimensional view
  • Source comparison bar charts
  • Calendar heatmaps
  • Correlation scatter plots

Voice Input

  • Uses Web Speech API
  • Real-time transcription
  • Ticker confirmation dialog

AI Insights

  • Automated trend analysis
  • Correlation calculations
  • Volatility assessment
  • Key topics extraction
  • Prediction indicators

🔧 Troubleshooting

Backend Issues

Problem: Module not found errors

cd backend
pip install -r requirements.txt --upgrade

Problem: FinBERT model download is slow

  • The first run will download the FinBERT model (~450MB)
  • Subsequent runs will use the cached model
  • Ensure stable internet connection

Frontend Issues

Problem: Dependencies not installed

rm -rf node_modules package-lock.json
npm install

Problem: Port 3000 already in use

  • Change port in vite.config.ts:
server: {
  port: 3001, // Use different port
}

Voice Input Issues

Problem: Voice input not working

  • Ensure you're using Chrome, Edge, or Safari
  • Allow microphone permissions
  • Check browser console for errors

🎓 Academic Project

This project was created for NLP Academic Evaluation and demonstrates:

  1. Natural Language Processing:

    • Sentiment analysis using FinBERT
    • Emotion classification
    • Text preprocessing

  2. Machine Learning:

    • Transfer learning with pre-trained models
    • Multi-class classification
    • Confidence scoring

  3. Data Visualization:

    • Interactive charts
    • Real-time updates
    • Responsive design

  4. Full-Stack Development:

    • RESTful API design
    • State management
    • Performance optimization

📝 License

Created for academic purposes. All rights reserved.

🙏 Acknowledgments

  • ProsusAI for FinBERT model
  • Hugging Face for Transformers library
  • News API for news data
  • Yahoo Finance for stock data

📧 Contact

For questions or issues, please contact the development team.

Made with ❤️ and AI

FinSentiment Pro - Empowering financial decisions with AI-powered sentiment analysis