Skip to content

Ragatanga

License: MIT Version

Ragatanga is a hybrid retrieval system that combines ontology-based reasoning with semantic search for powerful knowledge retrieval.

Features

  • 💪 Hybrid Retrieval: Combines SPARQL queries against an ontology with semantic search for comprehensive knowledge retrieval
  • 🧠 Adaptive Parameters: Dynamically adjusts retrieval parameters based on query complexity and type
  • 🔄 Multiple Embedding Providers: Support for OpenAI, HuggingFace, and Sentence Transformers embeddings
  • 💬 Multiple LLM Providers: Support for OpenAI, HuggingFace, Ollama, and Anthropic LLMs
  • 🌐 Comprehensive API: FastAPI endpoints for querying and managing knowledge
  • 📊 Confidence Scoring: Ranks results with confidence scores for higher quality answers
  • 🌍 Multilingual Support: Translates queries to match your ontology's language
  • ⚙️ Flexible Configuration: Comprehensive configuration options through environment variables and config module

Quick Start

# Install the latest version from PyPI
pip install ragatanga

import asyncio
from ragatanga.core.ontology import OntologyManager
from ragatanga.core.retrieval import AdaptiveRetriever

async def main():
    # Initialize with your ontology file
    ontology_manager = OntologyManager("path/to/ontology.ttl")
    await ontology_manager.load_and_materialize()

    # Create retriever
    retriever = AdaptiveRetriever(ontology_manager)

    # Query your knowledge base
    results = await retriever.retrieve("What is Ragatanga?")
    print(results)

if __name__ == "__main__":
    asyncio.run(main())

Documentation

Explore the documentation to learn more about Ragatanga: