The GraphRAG retrieval process uses a two-stage strategy: first, a high-recall one-hop graph traversal to identify candidate nodes, followed by a dense vector-based re-ranking step using OpenAI embeddings and cosine similarity to refine the results.

Authors

• Person: Not available Organization: arXiv
  Efficient Knowledge Graph Construction and Retrieval from ... - arXiv

Sources

• Efficient Knowledge Graph Construction and Retrieval from ... - arXiv arxiv.org arXiv via serper

Referenced by nodes (3)

• OpenAI entity
• embeddings concept
• cosine similarity concept