New Approaches to Diversity and Novelty in Recommender Systems

Beyond accuracy, novelty and diversity have attracted increasing interest as quality factors of Recommender Systems (RS) in the last few years. This paper presents work in progress towards the application of intentoriented IR diversity techniques to the RS field, and the formalization of novelty and diversity metrics for RS. Experimental results show that the resulting diversification techniques produce interesting results with respect to the metrics defined here.