Web使用挖掘在客户关系管理中的应用

Application in Customer Relationship Management with Web Usage Mining

Web使用挖掘作为一种实用的客户关系管理技术具有重要的应用价值,而如何挖掘具有单变量的客户访问时间序列是该方法的瓶颈。针对这一问题,提出基于编辑距离的动态聚类算法分析客户的访问序列,并为客户关系管理提供决策依据。首先对Web日志文件进行预处理;然后采用编辑距离的动态聚类进行分类,通过编辑距离将单变量的时间序列转化为聚类矩阵;最后结合实际应用进行说明。

As a applied technology in CRM,Web usage mining has important application value.But how to mine univariate time series is the bottleneck of this method.For the problem,the paper proposes an approach to group time series and provides decision support to CRM.Firstly,pretreating the WebLog.Then,classing the customer using the dynamic clustering algorithm based on edit distance.Finally,the paper analyses it by a case.