Decoder. plain-English AI glossary

Keyword Search

● Core

Also called Keyword Matching

The plain "match these words" search everyone already knows from the web.

Think of it like

Typing words into a search box and getting back pages that contain them.

Example

A user types "invoice pdf export" and gets docs mentioning those words, ranked by how strongly they match.

How it actually works

Keyword search ranks documents by how well their words match the query, typically via inverted indexes and frequency-based scoring. It is the familiar baseline: cheap, fast, and good when the user’s words match the content’s words. It falls down on synonyms, misspellings, and intent — which is why "just add keyword search" is rarely the whole answer anymore.

For product teams

A dependable floor for search quality and a sanity check against fancier methods; not sufficient alone for natural questions.

For engineers

Inverted-index term matching with frequency scoring; the lexical half of most hybrid stacks.

Related

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