基于关联规则的web浏览路径相似程度计算

目前,电子商务网站要想盈利,就必须真正了解自己的客户以便为他们提供更好的服务.将具有相似兴趣爱好的客户聚集成类,对大量客户进行细分,形成不同的客户群体,再有针对性地对他们提供服务是现在常用的技术.对客户进行聚类分析是建立在对用户浏览路径进行相似程度计算的基础上的.传统的基于网页距离计算的浏览路径相似程度分析存在着较大的误差.本文提出的基于关联规则的Web浏览路径相似程度分析方法较好地解决了这个问题,通过实验也验证了算法的正确性.

Association rule-based similarity analysis for web access path

Currendy, if e-commerce web stations want to get profit, they must understand their clients to serve them better. A common used technique is to subdivide the clients to several partitions by clustering them in terms of their interests and then to provide personal service for different clusters. Clustering analysis is based on calculating the similarity degree of the user access paths. But, traditional method for similarity degree analysis based on the distance of web pages bring about large errors. Thus, this article provided a new method which utilizes association rules to analyze similarity degree of the web user access paths. Experiments proved the correctness of the algorithm.