Vector Database
読み方：Vector Database
A database designed to store, index, and search high-dimensional vector embeddings. Enables semantic (meaning-based) similarity search rather than exact keyword matching. Essential infrastructure for RAG systems and AI-powered search. Popular options include Pinecone, Chroma, and Weaviate.
What is a Vector Database
A vector database stores data as numerical vectors (embeddings) and enables similarity search—finding items that are semantically similar even when they don't share exact keywords. This is the infrastructure that makes AI-powered search and RAG systems work.
How It Differs from Traditional Databases
| Aspect | Relational DB | Vector DB |
|--------|--------------|-----------|
| Search method | Exact/fuzzy match | Semantic similarity |
| Data format | Structured rows | High-dimensional vectors |
| Primary use | CRUD operations | AI search, recommendations |
How Embeddings Work
Text like "What is the refund policy?" is converted to a vector like [0.12, -0.34, 0.78, ...] by an embedding model. Similar meanings produce similar vectors. The database finds nearest-neighbor vectors using efficient algorithms (ANN).
Popular Vector Databases
- • Pinecone: Managed service, production-grade, simple API
- • Chroma: Open-source, easy local development
- • Weaviate: Hybrid search (vector + keyword), self-hosted option
Use Cases
RAG knowledge bases, product recommendation engines, semantic document search, image similarity search.