Mining large streams of user data for personalized recommendations

Abstract

The Netflix Prize put the spotlight on the use of data mining and machine learning methods for predicting user preferences. Many lessons came out of the competition. But since then, Recommender Systems have evolved. This evolution has been driven by the greater availability of different kinds of user data in industry and the interest that the area has drawn among the research community. The goal of this paper is to give an up-to-date overview of the use of data mining approaches for personalization and recommendation. Using Netflix personalization as a motivating use case, I will describe the use of different kinds of data and machine learning techniques. After introducing the traditional approaches to recommendation, I highlight some of the main lessons learned from the Netflix Prize. I then describe the use of recommendation and personalization techniques at Netflix. Finally, I pinpoint the most promising current research avenues and unsolved problems that deserve attention in this domain.