relevance feedback

Terms from Artificial Intelligence: humans at the heart of algorithms

The glossary is being gradually proof checked, but currently has many typos and misspellings.

Relevance feedback is used in document retreival, recommender systems and other forms of intelligent user interfaces to allow users to say how good automated suggestions are. This is then fed back into the algorithms to improve future suggestions. In some cases relevance feedback is explicit, for example, adding a star rating or thumbs up, in others it is implicit for example, if the user clicks through for more information about a suggested item. Epistemic interaction can be used to design interactions that increase this feedback.

Used in Chap. 17: page 263; Chap. 19: page 299; Chap. 21: page 332