Definition
The process of retrieving high-dimensional vector embeddings from a database that are mathematically closest to a query vector using distance metrics; it requires balancing computational speed (latency) against retrieval accuracy (recall).
Search based on semantic meaning and context rather than exact keyword or character matching.
"Finding the nearest neighbors in a high-dimensional star map."
- Vector Embeddings(Prerequisite)
- Cosine Similarity(Component)
- HNSW (Hierarchical Navigable Small World)(Algorithm)
- Approximate Nearest Neighbor (ANN)(Optimization Technique)
Conceptual Overview
The process of retrieving high-dimensional vector embeddings from a database that are mathematically closest to a query vector using distance metrics; it requires balancing computational speed (latency) against retrieval accuracy (recall).
Disambiguation
Search based on semantic meaning and context rather than exact keyword or character matching.
Visual Analog
Finding the nearest neighbors in a high-dimensional star map.