Search Intent
Also known as: query intent, user intent.
What the person behind a query is trying to achieve: find a place, learn something, or do something such as buy. Relevance depends on it.
Synthetic dataExamples on this page use synthetic data created for illustration.
Definition
Search intent is what the person behind a query is trying to achieve: to find a specific site, to learn about something, or to do something such as buy. The same words can carry different intents, so relevance depends on intent as well as on matching words.
In plain terms
A well-known classification by Andrei Broder (2002) distinguishes navigational queries (take me to a known place), informational queries (I want to learn) and transactional queries (I want to do or buy something). Later work refined it, but the lesson stands: two queries with similar words can need very different results.
For evaluation this matters twice. Judges need to know the intent when they grade a result (see Relevance Judgments), and a ranking tuned for one intent can be poor for another.
Example
| Query | Likely intent | A relevant result looks like |
|---|---|---|
| “acme login” | Navigational | The sign-in page of Acme |
| “how does a heat pump work” | Informational | An explanatory article |
| “buy heat pump 3kw” | Transactional | Product pages with prices |
| “jaguar” | Ambiguous | Depends: the animal, the car brand, or something else |
The last row shows why a system sometimes shows a mix, or asks which meaning is wanted.
Common misunderstandings
- “The same words always mean the same intent.” Context and the person decide.
- “Intent can always be read from the query.” Short queries are often genuinely ambiguous.
- “There are only three intents.” Three is a classic starting classification; real systems use more detailed ones.
References
- Andrei Broder. A Taxonomy of Web Search . ACM SIGIR Forum, 2002. DOI: 10.1145/792550.792552
- Bernard J. Jansen, Danielle L. Booth, Amanda Spink. Determining the Informational, Navigational, and Transactional Intent of Web Queries . Information Processing & Management, 2008. DOI: 10.1016/j.ipm.2007.07.015