摘要
对影响Web使用挖掘效果的会话识别方法进行理论研究,将会话识别按照对用户行为的不同假设分为基于时间的、基于导航的和基于语义的三种启发式方法,并对每种方法又进行细分研究,对会话识别理论方法进行综述,讨论这三种方法的各自优点和存在的问题。在对会话识别的方法进行综合比较的基础上,指出会话识别方法研究的两个趋势,一个是表示Web日志访问请求所代表的语义,一个是分析用户行为。
This essay is a theoretical research on the session identification approaches that will affect the effect of Web usage mining,and the session identification approaches are divided into three heuristics-based on time,navigation,and semantic.Moreover,each heuristic is divided and studied.The theoretical approaches are summarized,and their advantages,shortcomings and differences are discussed.By the end of this essay,the two possibilities of improving session identification approaches are provided.The one is the semantics of the request in web log,the other one is the analysis on user's behavior.
出处
《图书馆学研究》
CSSCI
北大核心
2012年第8期5-8,4,共5页
Research on Library Science
关键词
WEB使用挖掘
会话识别
语义会话识别
web usage mining session identification semantic session identification