RAG & Search
Sparse retrieval
Retrieval that uses mostly word-frequency or token-based representations.
Example
A sparse search emphasizes the uncommon words that appear in a query and document.
Why it matters
It offers an efficient way to find material using recognizable term patterns.
You'll hear it
“That uncommon word made the correct document easier to find.”
Can you control it?
Set by the team
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Does sparse mean the collection has few documents?
- No. It refers to a description with many unused positions, rather than the number of documents.
- Can it find differently worded expressions of the same idea?
- Sometimes, but exact term-based methods can miss matches without shared wording.
- Are common words always useless?
- No. Their importance depends on the question, language and search method.