用户行为模式挖掘问题的研究  

Research on User's Behavior Pattern Mining

作　　者：刘洪辉[1] 吴岳芬[2] 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074 [2]湖南理工学院计算机系,湖南岳阳414006

出　　处：《计算机技术与发展》2006年第5期50-52,55,共4页Computer Technology and Development

摘　　要：在软件可用性测试中,分析用户行为模式是一个关键的问题。为解决具有序列长度长、以序列片断为支持度计算依据等特点的用户行为模式挖掘问题,提出了一种有效的基于前缀树的频繁事件序列扩展方法,给出了比特图索引表的构造、事件扩展、事务扩展以及支持度计算的算法。使频繁事件序列能够简单快速地被确定。Analyzing a user＇s behavior pattern is a key problem in software usability testing. To solve user＇s behavior pattern mining prohlem, which is characterized as long sequence, calculation support according to sequence episode, an effective method based on extension of the frequent event sequence of the prefix tree are proposed, The algorithms for hit map index construction, event ext^ion, transaction extension and support calculation are also presented. Making frequent event sequence be sured quickly and simply.

关 键 词：事件序列 序列模式挖掘 用户行为模式挖掘 数据挖掘 软件可用性 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]