摘要
现有的静态 Web 站点结构不能满足人们准确地找到所需信息和享用个性化服务的要求。本文不但通过Web 日志文件的挖掘,找出用户的频繁访问路径来改进 Web 站点结构,而且分析当前访问页面与后续候选推荐页面的内容相关性,形成经过内容裁剪的个性化页面来压缩 Web 页面内容。这样,用户可快速定位到频繁访问的后续页面位置,且页面内容大多是用户感兴趣的主题信息。在此基础上,提出了一个自适应站点模型 AdaptiveSite,经过推荐质量分析,该模型具有较好的优化性能。
The current Web site is static, which can＇t meet people＇s requirements of finding useful information accurately and getting personalized services. This paper puts forward an optimal solution which not only improves web site structure by mining Web log files and finding users＇ frequently visited paths, but also forms content-pruned and personalized pages by analyzing the relevancy of thecontent of the latest visited page and that of the candidate recommended page. Thus, users can quickly locate the next frequently visited pages, which are almost interesting to users. Based on above algorithm, a model on adaptive Web site, called ＂AdaptiveSite＂, is presented. As it is proved by the recommendation quality analysis, this approach has a good performance of optimal recommendation.
出处
《计算机科学》
CSCD
北大核心
2006年第4期126-129,共4页
Computer Science
基金
国家自然科学基金资助(60205007)
广东省自然科学基金资助(001264
031558)
广东省科技计划项目资助(2003C50118)
广州市科技计划项目资助(200223-E0017)
