Real-time Cryptocurrency Sentiment Analysis Platform - A comprehensive solution that analyzes cryptocurrency news sentiment using dual AI models (VADER & FinBERT) across mobile and web platforms.

CryptoSentiment-Core is a sophisticated platform that provides real-time sentiment analysis for cryptocurrency news using two cutting-edge AI models. The system aggregates news from multiple sources, processes content through VADER (rule-based) and FinBERT (transformer-based) models, and delivers actionable insights through both mobile and web interfaces.

• Dual AI Model Approach: Leverage both speed and accuracy with VADER's fast processing and FinBERT's financial domain expertise
• Cross-Platform Accessibility: Access insights via mobile app (iOS/Android) or web dashboard
• Real-time Analysis: Live news monitoring with immediate sentiment processing
• Multi-source Integration: Aggregate data from leading cryptocurrency news sources
• Professional Design: Modern dark-themed interface with responsive design

The intuitive mobile interface showing real-time market sentiment

• Native Cross-Platform Experience: Seamless iOS and Android support
• Real-time Sentiment Feed: Live news pulse with color-coded sentiment indicators
• Offline Capability: Core functionality available without constant internet connectivity
• Intuitive Navigation: Bottom navigation bar for quick access to different sections
• Performance Optimized: Efficient data handling and smooth user experience

Comprehensive dashboard with market overview and sentiment analysis

• Sentiment Visualization: Interactive pie charts comparing positive, negative, and neutral sentiments
• Real-time News Feed: Curated cryptocurrency news with sentiment indicators
• Multi-coin Support: Analyze sentiment for Bitcoin, Ethereum, Ripple, and other major cryptocurrencies
• Advanced Filtering: Sort news by date, coin, sentiment, and source
• Performance Metrics: Historical sentiment trends and comparison analysis

Detailed sentiment analysis for various cryptocurrencies

• Per-coin Analysis: Individual sentiment statistics for popular cryptocurrencies
• Sentiment Comparison: VADER vs FinBERT results for validation
• Interactive Grid View: Cards for each coin with clear visual indicators
• Market Health Score: Aggregate market sentiment across all cryptocurrencies

Real-time live market monitoring system

• Live News Feed: Stream updates from major cryptocurrency sources
• Live Processing: Instant sentiment analysis using dual AI models
• Model Comparison: View how both VADER and FinBERT assess each piece of news
• Sources Diversity: Integrate feed from multiple platforms (Coindesk, CoinTelegraph, etc.)

Interactive AI model testing environment

• Model Comparison: Test both VADER and FinBERT on custom text
• Real-time Results: Instant sentiment analysis with confidence scores
• Sample Scenarios: Pre-loaded examples for quick testing
• Educational Tool: Understand how different models interpret sentiment

• Type: Rule-based sentiment analyzer
• Speed: Fast processing (~2ms per article)
• Strengths: Excellent for real-time processing, handles social media text well
• Use Case: Primary model for live news monitoring
• Implementation: NLTK-based with cryptocurrency-specific lexicon adjustments

• Type: Transformer-based deep learning model
• Accuracy: High precision for financial domain text
• Strengths: Contextual understanding, handles complex financial terminology
• Use Case: Validation and detailed analysis of important news
• Implementation: Pre-trained ProsusAI/finbert model fine-tuned for financial sentiment

• Speed vs Accuracy: VADER for quick processing, FinBERT for validation
• Cross-verification: Compare results for higher confidence
• Redundancy: System remains functional even if one model fails
• Performance Optimization: Use appropriate model for different use cases

┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│   Mobile App    │    │  Web Dashboard  │    │   Data Sources  │
│   (Flutter)     │    │    (Next.js)    │    │ (RSS/API feeds) │
└─────────┬───────┘    └─────────┬───────┘    └─────────┬───────┘
          │                      │                      │
          └──────────────────────┼──────────────────────┘
                                 │
                    ┌─────────────────────┐
                    │   Backend API       │
                    │   (FastAPI/Python)  │
                    └─────────────────────┘
                                 │
                    ┌─────────────────────┐
                    │   Database Layer    │
                    │   (SQLite)          │
                    └─────────────────────┘

• Framework: FastAPI with Uvicorn ASGI server
• Database: SQLite with SQLAlchemy ORM
• AI Libraries:
  • nltk for VADER sentiment analysis
  • transformers for FinBERT model
  • torch for PyTorch deep learning
• Data Sources:
  • CryptoCompare API
  • NewsAPI.org
  • RSS feeds (CoinDesk, CoinTelegraph, CryptoNews, etc.)
• Processing: Concurrent news fetching and sentiment analysis

• Framework: Next.js 16 with App Router
• Language: TypeScript
• Styling: Tailwind CSS with custom dark theme
• Data Visualization: Recharts for interactive charts
• State Management: React Context API
• HTTP Client: Axios for API communication

• Framework: Flutter 3.8+
• Language: Dart
• State Management: GetX pattern
• UI Components:
  • fl_chart for data visualization
  • http for API calls
  • url_launcher for external links
• Design: Material Design with custom dark theme

1. News Aggregation: Multiple RSS feeds and APIs
2. Content Processing: Text cleaning and preparation
3. Dual Analysis: Simultaneous VADER and FinBERT processing
4. Result Storage: Database with both model results
5. API Distribution: RESTful endpoints for client consumption
6. Client Display: Mobile and web interfaces

• News Table: Historical news articles with sentiment data
• Live News Table: Real-time news with dual model results
• Sentiment Fields: Separate columns for VADER and FinBERT results
• Metadata: Source, timestamp, URL, and processing information

• Python: 3.8 or higher
• Node.js: 18 or higher
• Flutter: 3.8 or higher
• Android Studio: For mobile development (optional)
• Git: For version control

# Python packages (automatically installed via requirements.txt)
fastapi
uvicorn
pandas
nltk
sqlalchemy
transformers
torch>=2.6
scipy
requests
feedparser
beautifulsoup4
lxml

# Navigate to backend directory
cd backend

# Install Python dependencies
pip install -r requirements.txt

# Start the FastAPI server
uvicorn main:app --reload

# Server will be available at http://127.0.0.1:8000

# Navigate to web dashboard directory
cd web_dashboard/crypto-dashboard

# Install Node.js dependencies
npm install

# Start development server
npm run dev

# Dashboard will be available at http://localhost:3000

# Navigate to mobile app directory
cd mobile_app/crypto_sentiment_app

# Install Flutter dependencies
flutter pub get

# Run on connected device or emulator
flutter run

# For web version (alternative)
flutter run -d chrome

Create a .env file in the backend directory:

# News API Configuration
NEWS_API_KEY=your_newsapi_key_here

# Database Configuration
DATABASE_URL=sqlite:///./crypto.db

# Server Configuration
HOST=127.0.0.1
PORT=8000

1. NewsAPI: Get free API key from newsapi.org
2. CryptoCompare: No API key required for basic usage
3. RSS Feeds: No configuration needed (public feeds)

• Development: Use --reload flag for auto-restart
• Production: Use process managers like gunicorn or pm2
• Database: SQLite for development, PostgreSQL for production

1. Access: Visit http://localhost:3000
2. Features:
   • Overall market sentiment visualization
   • Latest news feed with sentiment indicators
   • Coin-specific analysis (search by name)
   • Real-time data updates

1. Navigation: Click "Market" in sidebar
2. Features:
   • Per-coin sentiment cards
   • Comparison between cryptocurrencies
   • Detailed statistics view
   • Quick navigation to detailed analysis

1. Navigation: Click "Live Monitor" in sidebar
2. Features:
   • Real-time news feed
   • Live sentiment processing
   • Model comparison view
   • Manual sync capability
   • Adjustable news count

1. Navigation: Click "AI Playground" in sidebar
2. Features:
   • Text input for custom analysis
   • Model selection (VADER/FinBERT)
   • Real-time results with confidence scores
   • Sample scenarios for testing
   • Educational comparison tool

• Overview: Market sentiment summary
• Quick Stats: Positive/Negative/Neutral counts
• Recent News: Latest articles with sentiment
• Navigation: Bottom tab bar for sections

• Real-time Feed: Streaming news updates
• Sentiment Indicators: Color-coded sentiment badges
• Source Information: News source attribution
• External Links: Tap to read full articles

• Theme Toggle: Dark/Light mode switch
• AI Model Selection: Choose default processing model
• About Information: Version and project details

# Get news articles
GET /api/news?skip=0&limit=50&q=Bitcoin

# Get sentiment statistics
GET /api/stats?q=Bitcoin

# Get live news
GET /api/live_news?limit=20

# Analyze text with specific model
POST /api/analyze_text
{
  "text": "Bitcoin price surges after positive regulatory news",
  "model": "vader"
}

# Re-analyze database with specific model
POST /api/reanalyze_db
{
  "model": "finbert",
  "limit": 100
}

# VADER statistics
GET /api/vader_stats?q=Bitcoin

# FinBERT statistics
GET /api/finbert_stats?q=Bitcoin

# Live VADER stats
GET /api/live_vader_stats

# Live FinBERT stats
GET /api/live_finbert_stats

📱 Home Screen

Dashboard overview showing market sentiment and recent news

⚡ Live Market Pulse

Real-time news feed with dual sentiment analysis

🧪 AI Playground

Mobile interface for testing sentiment models

🖥️ Main Dashboard

Comprehensive market overview with sentiment visualization

📈 Market Analysis

Per-coin sentiment analysis with detailed statistics

📡 Live Monitor

Real-time news processing with model comparison

🔬 AI Playground

Interactive testing environment for sentiment models

• Backend: Follow PEP 8 Python standards
• Frontend: Use TypeScript with strict typing
• Mobile: Follow Flutter/Dart best practices
• Documentation: Keep README and code comments up to date

• main: Production-ready code
• develop: Active development branch
• feature/branch-name: New features
• hotfix/branch-name: Critical bug fixes

1. Fork the repository
2. Create feature branch from develop
3. Implement changes with proper testing
4. Update documentation if needed
5. Submit pull request to develop branch

• Backend: Unit tests for API endpoints and AI models
• Frontend: Component tests and integration tests
• Mobile: Widget tests and integration tests
• End-to-End: Cross-platform functionality testing

• Use GitHub Issues for bug reports
• Include detailed reproduction steps
• Specify platform (mobile/web/backend)
• Add relevant screenshots or logs

This project is licensed under the MIT License - see the LICENSE file for details.

• Free for personal and commercial use
• Attribution required in distributions
• No warranty provided

• FastAPI: High-performance Python web framework
• Next.js: React framework for production
• Flutter: Google's UI toolkit for mobile
• NLTK: Natural Language Toolkit for VADER
• Hugging Face Transformers: For FinBERT implementation

• CryptoCompare: Cryptocurrency news API
• NewsAPI: General news aggregation
• RSS Feeds: Various cryptocurrency news sources
• Coindesk, CoinTelegraph: Industry-leading publications

• ProsusAI: FinBERT model development
• NLTK Community: VADER sentiment analysis tools
• Open Source Community: Various libraries and tools

• GitHub Discussions: Project discussions and Q&A
• Issues Tracker: Bug reports and feature requests
• Documentation: Comprehensive guides and API references

• Enterprise Licensing: Contact for commercial support
• Custom Development: Integration and customization services
• Training: Team onboarding and best practices

Built with ❤️ for the cryptocurrency community

Back to Top • GitHub Repository • Live Demo