Adaptive multi-attribute diversity for recommender systems

Providing very accurate recommendations to end users has been nowadays recognized to be just one of the tasks an effective recommender system should accomplish. While predicting relevant suggestions, attention needs to be paid also to their diversification in order to avoid monotony in the returned list of recommendations. In this paper we focus on modeling user propensity toward selecting diverse items, where diversity is computed by means of content-based item attributes. We then exploit such modeling to present a novel approach to re-arrange the list of Top-N items predicted by a recommendation algorithm, with the aim of fostering diversity in the final ranking. An extensive experimental evaluation proves the effectiveness of the proposed approach as well as its ability to improve also novelty and catalog coverage values.