基于加权关联规则的用户关注项目推荐算法  

User Attention Items Recommendation Algorithm Based on Weighted Association Rules

作　　者：陈华月[1] 余刚[1] 朱征宇[1] 

机构地区：[1]重庆大学计算机学院

出　　处：《计算机工程》2006年第6期86-88,共3页Computer Engineering

基　　金：重庆大学骨干教师基金资助项目(2003A33)

摘　　要：根据每个项受关注的程度对其赋予权值,体现每个项在项集中的重要性,引入选择关注度,体现用户选择项时的倾向,推荐时将其与关联规则的置信度相结合,从而为用户更准确的提供关注项目推荐。在实验中比较了文中的算法和一种基于最近邻用户协作过滤算法,实验验证了文中的算法在为用户进行推荐项目时用户的点击率和算法的性能都有所提高。According to each item of degree that is concerned by users, this paper gives the weight, concentrating importance of each item. It imports the choice attention and explains the trend of items they are selected in order to provide the items recommendation more accurately for the users. In experiment, the method in this paper and a kind of according to most the close neighbors is compared, to make to recommend the method. Experiment shows that the algorithm has better performance and accuracy.

关 键 词：加权关联规则 选择关注度 项目推荐 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]