Yosef-Ali/amharic-dataset-mcp

# Amharic Dataset MCP Tools 🇪🇹 **Production-ready MCP (Model Context Protocol) tools for authentic Amharic dataset collection, enhancement, and quality scoring.** ## 🚀 Features ### 📰 Authentic Data Collection - **Ethiopian news sources**: BBC Amharic, VOA Amharic, Ethiopian Reporter - **Social media integration**: Facebook groups, Telegram channels - **Literature sources**: Ethiopian books, religious texts, educational materials - **Smart Amharic detection**: Unicode-based authentic text filtering ### 🔮 RAG-Based Enhancement - **Context-aware corrections**: Uses high-quality Amharic knowledge base - **Vector similarity search**: FAISS-powered intelligent matching - **Grammar pattern fixes**: Natural expression improvements - **Cultural authenticity**: Ethiopian context validation ### ⚡ Multi-Dimensional Quality Scoring - **Grammar quality**: Pattern-based validation (30% weight) - **Amharic purity**: Unicode character analysis (25% weight) - **Cultural authenticity**: Ethiopian keyword density (20% weight) - **Conversation naturalness**: Question-answer patterns (15% weight) - **Vocabulary richness**: Word diversity metrics (10% weight) ### 🗄️ Database Integration - **Multi-database support**: SQLite, PostgreSQL, MySQL - **Structured storage**: Metadata, quality scores, timestamps - **Fast retrieval**: Indexed searches for training data - **Batch processing**: Scalable dataset operations ## 📦 Installation ```bash # Clone repository git clone github.com cd amharic-dataset-mcp # Install package pip install -e . # Install with development tools pip install -e ".[dev]" # Install with GPU support pip install -e ".[gpu]" # For Gemini integration pip install google-generativeai # For Qwen models pip install transformers torch # Complete installation with all AI models pip install -e ".[dev,gpu]" google-generativeai transformers torch ``` ## 🔧 Quick Start ### 1. Start MCP Server ```bash # Start the Amharic dataset M …