应用个性化推荐的Web日志关联规则挖掘算法研究  

Reserch of Web Log Associated Rules Mining Algorithm Applied to Web Personalized Recommendation

作　　者：邹丽霞[1] 

机构地区：[1]河南广播电视大学,郑州450008

出　　处：《河南科学》2010年第9期1125-1129,共5页Henan Science

基　　金：河南省科学技术厅科技发展计划项目(092300410040)

摘　　要：对传统的关联规则挖掘算法FP-Growth方法进行改进,提出FP-Mine算法,并应用该算法对Web日志进行挖掘,探寻用户访问站点页面之间的关联规则,来帮助管理员改善站点的设计和企业改进市场商务决策.实验结果证明FP-Mine算法在生成频繁项集及关联规则的过程中,只需存储i-size和(i+1)-size频繁项集的节点的Freq-Set-Tree,且立即在其之上生成规则,所以缩短规则生成的时间,提高规则生成效率,同时释放i-size项集的节点,有效地节省内存空间.This paper proposes the FP-Mine algorithm which is improved according FP-Growth and uses it to mine the Web logs.The new algorithm can find the associated rules in the webs,then helps web managers or companies to improve the web designs or business decisions.The experiments show that in the process of using FP-Mine algorithm to find frequent itemsets and associated rules,only the Freq-Set-Tree of i-size and（i＋1）-size frequent itemsets nodes and the rules can be get in it.So FP-Mine is more efficient in time and space.

关 键 词：WEB日志挖掘 关联规则 加权支持度 加权置信度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]