Vector database
Vector database: A database designed to store embeddings and retrieve similar items through vector indexes. It combines approximate nearest-neighbor search with metadata, persistence and filtering. Metric, index and update strategy affect recall, latency and cost more than the product name.
In italiano: Vector DatabaseLegge I · Embedding, feature e spazi vettorialiIt combines approximate nearest-neighbor search with metadata, persistence and filtering. Metric, index and update strategy affect recall, latency and cost more than the product name.
Examples
- Semantic document search
- Similarity recommendations
- Content deduplication