Towards Diverse Recommendation

Novelty and Diversity in Recommender Systems 2011.

计算机科学;

Neil J. Hurley

Others  :  http://ceur-ws.org/Vol-816/keynote.pdf
PID  :  42543

学科分类：计算机科学（综合）

来源: CEUR

【 摘 要 】

In recent years great strides have been made in improving the accuracy of recommender systems from the point-of-view of their ability topredict users’ ratings for unrated content given a database of past ratings. In a context where the system should ultimately recommend a listof items to the end-user, such accurate rating predictions can be seen as just one possible input into the decision system that selects therecommended content. It has been recognized for several years now that other qualities of therecommended list are also important in thisselection process; it is not simply a matter of recommending those items with highest predicted ratings. In particular, a good system shouldoffer a diverse choice of relevant items, allowing users to select from across their broad range of tastes. It is worth emphasizing thatdiversifying the recommendation is not simply a matter of selecting a set of highly dissimilar items for recommendation, since relevance isstill a primary concern – increasing diversity while maintaining system performance, as measured by a relevance metric is a significantchallenge. Research in diverse recommendation is still in an early stage; while a number of algorithms and systems for diverserecommendation have been proposed, many different performance measures and evaluation methodologies are being used making itdifficult to compare across different approaches. In this talk, I attempt to summarize the state-of-the-art in diverse recommendation,bringing together the different approaches that have been proposed in recent years and the various performance measures that have beenused. The goal is to set the context and to propose some ideas to generate what should be some interesting and controversial discussions during the remainder of the workshop.