摘要
协作过滤是当今应用较为成功的个性化服务技术,Web日志可以为个性化服务技术提供重要的数据源,只要对日志数据进行高效预处理,就能提高协作过滤算法有效性和个性化服务质量。结合实际日志数据的处理,给出了基于协作过滤的Web日志数据预处理过程结构图和一种可行的数据预处理方法,该方法不仅可以提供更加干净、规则的数据源,而且在用户兴趣度量方面,弥补了以往诸多兴趣度量方法的不足,为协作过滤算法提供了更加准确的数据支持。
Collaborative filtering is one of the most successful applications in the field of personalized service. The Web log can provide personalized service technology with important data resource. And the validity of collaborative filtering algorithm and the quality of personalized service will be improved as long as the effective preprocessing to log data has been done. On the foundation of dealing with material log data, it gives a structure figure and a method of data preprocessing based on collaborative filtering. This method can not only provide cleaner and more regular data source but also make up the defects of former interests measurement in the aspect of user＇s interests measurement, which provides a more precious sustainment for collaborative filtering algorithm.
出处
《重庆邮电学院学报（自然科学版）》
2006年第5期646-649,共4页
Journal of Chongqing University of Posts and Telecommunications(Natural Sciences Edition)
基金
重庆市自然科学基金重点资助项目(2005BA2003)
关键词
个性化服务
协作过滤
WEB日志
数据预处理
用户兴趣
personalized service
collaborative filtering
Web log
data preprocessing
user＇s interests
