kara-toolkit 0.1.0

Knowledge-Aware Re-embedding Algorithm - Efficient RAG knowledge base updates

KARA - Knowledge-Aware Re-embedding Algorithm

KARA is a Python library for efficient document updates in RAG systems. It minimizes embedding operations by intelligently reusing existing chunks when documents are updated.

Installation

pip install kara-toolkit

Quick Start

from kara import KARAUpdater, RecursiveCharacterChunker

# Initialize
updater = KARAUpdater(
    chunker=RecursiveCharacterChunker(chunk_size=1000),
    epsilon=0.1
)

# Process initial documents
updater.initialize(["Your document content..."])

# Update with new content
result = updater.update(["Updated document content..."])
print(f"Efficiency: {result.efficiency_ratio:.1%}")

How It Works

KARA formulates the chunking problem as a DAG (Directed Acyclic Graph) for a single document where each node represents a position in the document splits, and edges represent possible chunks. It then uses Dijkstra's algorithm to find the optimal chunking path.

Examples

See the examples/ directory for more usage examples.

License

CC BY 4.0 License - see LICENSE file for details.