A New Collaborative Filtering Algorithm for Recommender Systems
Kendall correlation based collaborative filtering algorithm was proposed for the recommendation problem. The Kendall correlation method is used to measure the correlation amongst users, which considers only the relative order of the ratings. Although the experiments show that Kendall based algorithm performs slightly worse than the classic Pearson algorithm, it is based on a more general model and thus could be more widely applied in e-commerce. Another discovery is that the consideration of only positive correlated neighbors in prediction, in both Pearson and Kendall algorithms, achieves higher accuracy than the consideration of all neighbors, with only a small loss of coverage.
Kendall correlation collaborative filtering recommender system positive correlation data mining
Yao Yu Shanfeng Zhu Jinshuo Liu Xinmeng Chen
Computer School, Wuhan University, Wuhan Hubei 430072, China Institute for Chemical Research, Kyoto University, Uji Kyoto 611-0011, Japan
国际会议
杭州
英文
634-636
2006-10-12（万方平台首次上网日期，不代表论文的发表时间）