« Back to Glossary Index

A recommender system is a solution which provides suggestions for items that are most pertinent to a particular user. They are particularly useful when there is an overwhelming number of items from which an individual can choose.
The two main approaches for designing a recommender system are «collaborative filtering» and «content-based filtering»

  • Collaborative filtering is a method of making predictions about the interests of a user by collecting preferences from many users, all of them similar to the user who needs recommendations. Hence the underlying assumption of such approach is that if a user A has the same opinion of user B on an issue, A is more likely to have B’s opinion on a different issue.
  • Content-based filtering methods are based on the analysis of features and characteristics of the item and a profile of the user’s preferences, therefore they consider recommendation as a classification problem tailored for the specific user, classifying the user’s likes and dislikes according to the features of each item.