Fuzzy logic in recommender systems

Abstract

A recommender system studies the past behaviour of a user and recommends relevant and accurate items for the user from a large pool of information. For user ‘u’ a recommender system filters relevant and accurate information by finding items which are similar to the target items (items searched by user) and by finding (other) similar users which co-relate to the user ‘u’ interests and needs. For exhibiting this filtration, a recommender system uses the features of the items and maintains user profile which contains his past purchases, his buying pattern. These features and user profile imprecise, uncertainty and vague thus should be analysed carefully for optimal prediction. Fuzzy logic has been extensively used in the design of a recommender system to handle the uncertainty, impreciseness and vagueness in item features and user’s behaviour. In this chapter, the use of fuzzy logic in recommender system as well as the analytical framework for analysis of the design of a recommendation system is discussed. It also presents the analysis of the growth of fuzzy logic in recommender system and its applications.