Information retrieval
Data structuring
The shape of the data before a ranker runs. Vector search and lexical tagging sit under this subject. The structure decides what close is allowed to mean.
Broader subject: Relevance
Within this subject
- Vector search — Search by nearest neighbor in an embedding space. I am studying when a dense signal helps, and when a precise lexical handle should outrank it.
- Lexical tagging — Tokens, entities, and tags taken from the text itself, including tags extracted with classical tools or with a language model. A label a person can read, used as a handle for lexical retrieval.