Latent features based prediction on new users' tastes

Ming Chu Chen, Hung-Yu Kao

研究成果: Conference article同行評審


Recommendation systems have become popular in recent years. A key challenge in such systems is how to effectively characterize new users' tastes - an issue that is generally known as the cold-start problem. New users judge the system by the ability to immediately provide them with what they consider interesting. A general method for solving the cold-start problem is to elicit information about new users by having them provide answers to interview questions. In this paper, we present Matrix Factorization K-Means (MFK), a novel method to solve the problem of interview question construction. MFK first learns the latent features of the user and the item through observed rating data and then determines the best interview questions based on the clusters of latent features. We can determine similar groups of users after obtaining the responses to the interview questions. Such recommendation systems can indicate new users' tastes according to their responses to the interview questions. In our experiments, we evaluate our methods using a public dataset for recommendations. The results show that our method leads to better performance than other baselines.

頁(從 - 到)176-187
期刊Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
8443 LNAI
發行號PART 1
出版狀態Published - 2014 一月 1
事件18th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2014 - Tainan, Taiwan
持續時間: 2014 五月 132014 五月 16

All Science Journal Classification (ASJC) codes

  • 理論電腦科學
  • 電腦科學(全部)


深入研究「Latent features based prediction on new users' tastes」主題。共同形成了獨特的指紋。