Recommendation systems are AI systems that analyse user behaviour and suggest useful content. Collaborative filtering: "People similar to you also liked this" — user-user or item-item similarity. Content-based filtering: "Films similar to what you've watched" — based on object characteristics.
Matrix factorisation (SVD, ALS) is the mathematical foundation of collaborative filtering. The cold start problem: insufficient data for a new user or new item. Hybrid systems combine both approaches. YouTube's recommendation system is one of the world's most attention-capturing — the filter bubble effect originates there.
Our courses on this topic