Glossary Entry

Collaborative Filtering

A recommendation approach that predicts preferences from patterns in many users' interactions rather than item content alone.

Recommenders Personalization

Seed source: Google ML Glossary

Collaborative filtering works by finding structure in user-item behavior, such as clicks, ratings, purchases, or watches. The system learns that users with similar histories may prefer similar items, even when the item content itself is limited.

This matters in the recommender posts because it contrasts neatly with content-based methods. Collaborative filtering can be very effective once enough interaction data exists, but it is weaker in cold-start settings.