Vector Database
A database optimised for storing and querying high-dimensional vector data.
Definition
A vector database is a specialised database designed to store and efficiently query high-dimensional vector data, such as text embeddings. Unlike traditional databases that match exact values, vector databases find similar vectors using distance metrics. This enables semantic search applications where you want to find content similar in meaning to a query, even without exact keyword matches.
More in Data Infrastructure
Chunking
Splitting documents into smaller segments for processing and retrieval.
Embedding
A numerical representation of text that captures its semantic meaning.
Knowledge Graph
A structured representation of entities and their relationships.
Reranking
Reordering search results using a more sophisticated model to improve relevance.
See Vector in action
Understanding the terminology is the first step. See how Conductor applies these concepts to solve real document intelligence challenges.
Request a demo