协同工作环境中基于用户行为的页面推荐  

Page Recommendation Based on User Behavior in Collaborative Working Environment

作　　者：王强[1] 谢建军[1] 李华飙[1] 南凯[1] 

机构地区：[1]中国科学院计算机网络信息中心协同工作环境研究中心,北京100190

出　　处：《计算机工程》2012年第3期249-251,共3页Computer Engineering

基　　金：国家"863"计划基金资助项目(2006AA01A120);中国科学院知识创新工程基金资助重大项目(KGGX1-YW-13)

摘　　要：Wiki被广泛应用于协同工作环境,能实现用户的协同写作,但由于Wiki页面间导航的缺失,使用户很难找到目标页面。为此,使用协同过滤技术,通过对用户历史行为的分析,建立页面之间的关联,实现页面推荐系统。该系统使用基于用户行为的隐性评分,解决了Wiki中显性评分缺失的问题,并针对Wiki中用户关系随时间变化的特点,设计基于关联规则的相似度函数。应用结果表明,该系统能帮助用户快速找到目标页面,提高用户的浏览效率。Wiki based collaboration platforms are used broadly, which brings a great of flexibility and effectiveness. However the lack of navigation supporting in Wiki makes it hard for user to find target page. By using collaboration ＇filtering technical, this paper implements a page recommendation system to build the connection of pages. The absence of explicit rating in Wiki is solved by using implicit rating extracting from user history behavior. For the relation of users is time related, a new similarity function is proposed. Application result shows that the page recommendation system enhances the user browsing experience in Wiki, and helps user find target page effectively.

关 键 词：协同过滤 页面推荐 隐性评分 协同工作环境 关联规则 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]