基于免疫聚类的Web日志挖掘

Web日志挖掘旨在使用数据挖掘技术从Web服务器日志文件中挖掘出有用的规律和模式,以此改进网站结构以及实现Web个性化服务。本文提出基于免疫聚类的Web日志挖掘算法,利用人工免疫系统的基本原理来进行用户聚类分析,从而发现相似客户群体、挖掘潜在客户。免疫聚类通过模拟免疫系统体液免疫应答的基本过程,提取出数据的基本特征,以此概括数据的分布特征,从而实现Web日志数据的无监督自组织聚类。通过在真实数据集上的实验证明了该算法的可行性和有效性。

Web Log Mining Based Upon Immune Clustering

Web log mining aims at mining some useful rules and patterns from web server log fde by means of data mining technology so that website structure can be improved and web personality service can be realized.An immune clustering algo- rithm of web log mining is presented in this paper.It applies artificial immune system to clustering analysis for the sake of finding similar customer population and mining underlying customer.Immune clustering simulates sap of immune respouse process of immune system and extracts the essential feature of data and generalize distribution feature of data,accordingly un- supervised clustering of web log mining can be realized.Experimental results applied on real data set show that immune clus- tering algorithm of web log mining is feasible and effective.