基于Web结构分区的协同过滤推荐算法研究  

Collaborative Filtering Recommendation Algorithm Research Based on Web Blocks

作　　者：邵延振[1] 蒙韧[2] 袁鼎荣[1] 李新友[1] 

机构地区：[1]广西师范大学计信学院,广西桂林541004 [2]广西师范大学财务处,广西桂林541004

出　　处：《计算机技术与发展》2010年第6期67-69,73,共4页Computer Technology and Development

基　　金：广西自然科学基金(桂科自0640069)

摘　　要：Intenet的快速增长导致了个性化服务的需求急剧增加。基于页面结构的信息提取与推荐是Web数据挖掘中三大研究领域之一。该研究的关键技术是识别Web页面的组织形式,从中挖掘所需要的个性化页面信息。基于Web数据挖掘的个性化信息推荐系统可以满足互联网未来发展趋势的需要。与传统的以页面为单位的Web信息提取相比,基于页面结构分区的信息推荐更符合实际情况,粒度优势明显。以一组数据为实例阐述了基于Web挖掘的协同过滤推荐算法是如何进行数据表示、近邻查询以及产生推荐页面分区信息的。With development of Internet,personalized service demand rapidly increased.Information extraction and recommendation based web page structure is one of three web data mining＇s research fields.Key technology of the research is how to recognize web page＇s organization form and mine the needed information.Personalized recommender system based web data mining can meet the need of the Internet＇s future development trends.Compared with based on web page,the based web block is more accordant to the fact and the advantage of granularity is evident.In this paper,with one set of data as an example collaborative filtering recommendation algorithm was elaborated based on web data mining how to work in the progress of data expression,neighbors queries and recommend generation.

关 键 词：WEB数据挖掘 推荐系统 协同过滤 页面分区 个性化信息 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]