Semantic Search
Search that matches on meaning rather than exact keywords, using embeddings to find passages whose sense is close to the query.
Semantic search is how agents find your content by meaning; writing clear, entity-rich, self-contained passages is what makes you retrievable when the words don't match.
- term
- Semantic Search
- category
- optimization
- short_def
- Search that matches on meaning rather than exact keywords, using embeddings to find passages whose sense is close to the query.
- long_def
- Semantic search embeds both the query and the corpus into a vector space and retrieves by similarity, so 'how do I make my site readable to bots' can match a page titled 'agent-readiness' with no shared keywords. It is the retrieval backbone of RAG and the reason meaning-rich, well-structured content is found even without literal keyword overlap.
- see_also
embeddingsvector-databaserag- etymology_origin
- — verify-against-primary-at-build ↗ https://en.wikipedia.org/wiki/Semantic_search — established IR term predating LLMs; embedding-based revival is recent
- related_to
embeddingsvector-databaseragreranking- contrast_with
- Unlike lexical (keyword) search that matches exact strings, semantic search matches meaning via embeddings — it can retrieve a relevant page that shares no words with the query.
- example
- A query for 'stop AI bots scraping me' semantically matches a page about robots.txt and pay-per-crawl even though it never uses the word 'scraping'.
- source
- https://en.wikipedia.org/wiki/Semantic_search
- status
- active
- why_it_matters
- Semantic search is how agents find your content by meaning; writing clear, entity-rich, self-contained passages is what makes you retrievable when the words don't match.
- sameAs
https://en.wikipedia.org/wiki/Semantic_search- bridge_entity
- rag
- last_verified
- 2026-07-06
- md_twin
- /glossary/semantic-search.md
last verified · by Özden Erdinc