Proceedings of the 2nd KDD Workshop on Large-Scale Recommender Systems and the Netflix Prize Competition

Abstract

In October 2006, Netflix released a large dataset and launched the Netflix Prize Competition that attracted over 20,000 participants from 167 countries. As a part of this competition, many interesting data mining recommendation techniques have been developed that, by the very nature of the Netflix dataset, should be scalable to large recommendation problems. This workshop follows on the previously held KDD 2007 workshop with the goal to exchange the ideas and present and discuss novel recommendation methods that specifically focus on the scalability and performance issues pertaining to large-scale datasets for recommender systems and the Netflix competition in particular. The call for papers for the workshop attracted 10 submissions, out of which 5 papers were selected for the presentation at the workshop among which 3 papers were by the authors of leading teams in the Netflix Prize competition. The workshop is opened with the keynote address on Exploring User Opinions in Recommender Systems by Bing Liu from the University of Illinois who provides his views on future developments of recommender systems based on opinion mining and sentiment analysis and sets a stage for the subsequent presentations and discussions among the participants.