A WEB RECOMMENDATION METHOD BASED ON USER INTEREST MODEL

摘要：

A method of Web recommendation based on user interest model is proposed from the point of mixed recommendation strategy in this paper. For analyzing features of the user interest model, recommendation satisfaction rate under different model parameter C was studied and effectiveness evaluation was done for the proposed method, UAP (by-user-average collaborative filtering predictor) and IAP (by-item-average collaborative filtering predictor) by metric of average absolute deviation (AAD). The optimal C for this interest model is inferred by experimental results which also indicate that the proposed recommendation method performs better than the other two in this research and the problem of data scarcity can be relieved to some extent.

关键词： Web recommendation interest model

作者: ZHEN-GANG YANG FEI-QI DENG

作者单位: Institute of Systems Engineering, South China University of Technology, Guangzhou 510640,China;Colle Institute of Systems Engineering, South China University of Technology, Guangzhou 510640,China

会议类型: 国际会议

会议名称: 2006 International Conference on Machine Learning and Cybernetics(IEEE第五届机器学习与控制论坛)

会议地点: 大连

会议语种:英文

页码: 4536-4539

在线出版日期: 2006-08-13（万方平台首次上网日期，不代表论文的发表时间）