Semantic search
Search that finds and ranks results by their semantic relevance to the query.
How is related content found?
Semantic search finds and ranks results by their semantic relevance to the query. In vector-based approaches, queries and source texts are represented as embeddings so that similar representations can be retrieved. Whether short questions match longer passages also depends on the model and content preparation. Search initially returns candidates from a collection. Generating an answer from them is a further processing step.
In a fictional knowledge portal, someone asks why a machine runs roughly, while a manual uses the wording vibrations in the drive. Suitable semantic search could find that passage despite the different wording. The correct machine, valid document version and permitted access still need checking. Similarity to a paragraph does not establish a particular fault cause. A planned assessment uses concrete queries and passages judged relevant for the task to investigate missing and unsuitable results.
Sources
Related terms
- EmbeddingAn embedding represents content as a numeric vector. The model and comparison method determine which similarities become visible.
- Vector databaseA vector database stores and searches vector representations. A similar match is not automatically a valid or permitted answer source.
- Retrieval-augmented generationRAG combines information retrieval with generated answers. Access to sources does not remove the need for quality and permission design.
- GroundingGrounding links an AI answer to verifiable evidence. A source link alone does not show that every claim is supported.
- Hybrid searchCombining lexical search and vector search into a shared result list.
- Document chunkingSplitting a document into smaller coherent sections for search and model context.
Your project with ASES
Describe your project to us. We will get back to you with an initial assessment.