Recommender systems are information filtering systems designed to generate accurate and relevant item suggestions for users based on available data. Netflix, Amazon, YouTube, and Spotify are some popular services with recommender systems in commercial use today.
There are three primary types of recommender systems:
- Collaborative filtering: Item recommendations reflect personalized preferences based on similarity to other users. Preferences can be explicit (item ratings) or implicit (item ratings per user-item interactions such as views, purchases, and so on).
- Content-based filtering: Item recommendations reflect contextual factors such as item attributes or user demographics; item suggestions can also use temporal factors such as location, date, and time where applicable.
- Hybrid: Item recommendations combine a variety (ensemble) of collaborative...