BM25 and Sparse Retrieval — Roadmap Step & Resources
Classic keyword search techniques and when they outperform embeddings
- Level: beginner
- Estimated time: 1 days
- Roadmap: Retrieval-Augmented Generation (RAG) & Vector Databases
Before this step
Study resources
- Hugging Face Documentation (Article) — Official documentation covering embedding models and retrieval-related libraries. Background reading/viewing for: BM25 and Sparse Retrieval.
- Pinecone Learning Center (Article) — A vector database provider's educational content on hybrid search and re-ranking. Relevant to: BM25 and Sparse Retrieval.
Part of
- Hybrid Search and Re-ranking — section
- Retrieval-Augmented Generation (RAG) & Vector Databases roadmap — the full learning path