Recommendation systems provide us a promising approach to deal with the information overload problem. Collaborative filtering is the key technology in these systems. In the past decades, model-based and memory-based methods have been the main research areas of collaborative filtering. Empirically, model-based methods may achieve higher prediction accuracy than memory-based methods. On the other side, memory-based methods (e.g. slope one algorithm) provide a concise and intuitive justification for the computed predictions. In order to take advantages of both model-based and memory-based methods, we propose a new approach by introducing the idea of machine learning to slope one algorithm. Several strategies are presented in this paper to catch this goal. Experiments on the MovieLens dataset show that our approach achieves great improvement of prediction accuracy. © 2012 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Wang, Y., Yin, L., Cheng, B., & Yu, Y. (2012). Learning to recommend based on slope one strategy. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7235 LNCS, pp. 537–544). https://doi.org/10.1007/978-3-642-29253-8_47
Mendeley helps you to discover research relevant for your work.