NPLWAP:一种新的Web序列模式挖掘算法  

Mining Web Access Sequence with a Novel Algorithm—NPLWAP

作　　者：林维仲[1] 张东站[1] 

机构地区：[1]厦门大学信息科学与技术学院,福建厦门361005

出　　处：《厦门大学学报（自然科学版）》2013年第1期25-31,共7页Journal of Xiamen University：Natural Science

基　　金：国家自然科学基金项目(50604012)

摘　　要：Web序列模式挖掘是将数据挖掘技术应用于Web访问序列,通过对Web访问序列的模式挖掘可以发现用户与网站交互的频繁模式,利用这些模式可以建模并分析用户与网站交互的模型,进而预测未来的访问模式,这对于构建智能化Web站点和开展电子商务活动有非常重要的意义.介绍了传统的PLWAP(position coded preorder linked WAPtree)算法,并在此基础上提出了一种对PLWAP算法中Header table的新的构建方法的改进算法(NPLWAP).在NPL-WAP算法中Header table的构建过程中每一步都只基于当前处理的节点的后缀树集,且Header table并不存储所有的后缀树集节点,而是只存储后缀树集根节点,从而减少挖掘过程的相关判断.通过对真实数据的实验对比可以看出NPLWAP算法在运行时间上比传统的PLWAP算法有了很大的改进.Web sequence pattern mining is an application of data mining on web access sequences. Mining web sequence patterns can be used to find the frequent patterns fr0m the interaction between user and the site,with these patterns we can model and analyze the interactive model and predict the future access model significantly benefiting intelligent web site construction and e-commerce busi- ness campaign. A traditional method-PLWAP is first discussed in this paper,and then we propose a new algorithm-NPLWAP-with a new method to construct Header table aiming to improve the mining process. In this algorithm we construct Header table based on the suffix trees under the current node each step,and save the root nodes only. From the result we obtain using NPLWAP on the real data we can infer it has better performance than PLWAP in terms of running time.

关 键 词：WEB数据挖掘 序列模式 PLWAP-tree NPLWAP 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]