基于web日志的连续频繁路径挖掘算法  

Algorithm of Mining Continuous Frequent Paths Based on Web Logs

作　　者：任永功[1] 付玉[1] 张亮[1] 

机构地区：[1]辽宁师范大学计算机与信息技术学院,辽宁大连116029

出　　处：《小型微型计算机系统》2008年第12期2272-2276,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(60603047);辽宁省教育厅高等学校科研基金(2008341);大连市优秀青年科技人才基金(2008J23JH026);辽宁省自然科学基金

摘　　要：频繁模式挖掘已成为web使用挖掘的研究热点,本文基于web日志提出一种新的频繁路径的挖掘算法.首先以线性回归方法求解兴趣度,其次将此兴趣度和页面名称作为最基本要素,建立的web浏览树,此浏览树可以完整地表现出web日志中连续、重复的浏览路径,最后在web浏览树上进行分析挖掘频繁浏览路径.该算法经实验证明能更全面地反映用户兴趣所在,挖掘的频繁浏览路径准确、合理.Freuent pattern mining is a focus on researching of Web Usage Mining . Based on web logs, this article proposes a new mining algorithm of the frequent paths. This algorithm first solves interest by the linear regression method ,then takes it and page name as the basic element, establishes web browsing tree which can display completely continual and iterative browsing paths in web logs ,finally carries on analysing and mining frequent browsing paths on the web browsing tree. The experiment proved that the algorithm can be more fully reflect the interest of users, the frequent paths are exact and reasonable.

关 键 词：WEB日志 浏览兴趣度 多元线性回归 频繁浏览路径 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]