Inverted Index
Also called Postings List
A word-to-documents map that makes keyword search fast — the backbone of classic search engines.
Think of it like
A book index listing, for each word, every page it appears on.
Example
Elasticsearch stores, for the word "refund", the list of every document ID containing it, so lookups are near-instant.
How it actually works
Instead of storing "document → words", an inverted index stores "word → documents", with positions and frequencies. To answer a query you intersect the lists for each term and score by relevance, typically BM25. It is decades-old technology and still unbeatable for exact-term retrieval, which is why hybrid systems keep it beside their shiny vector index.
For product teams
Why keyword search scales to billions of docs; the reliable half of a hybrid stack.
For engineers
Term → postings mapping with positions/frequencies; list intersection plus BM25 scoring drives lexical retrieval.
Related
- Lexical Search — The engine behind lexical search.
- Index — A specific kind of index.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome