一种混合粒子群优化模型的Web聚类方法*

通过分析在电子商务环境下Web挖掘的现状，考虑到Web数据的海量性和高维度性对抽取隐含的、事先未知的知识所带来的复杂性和维数灾，在普通K均值聚类、PSO聚类和K均值与PSO混合聚类算法的基础上，提出了一种将主成分分析与PSO混合聚类算法相结合的模型来对Web服务器中的日志文件进行聚类分析，将抽取的相关Web数据进行主成分分析，分析结果作为PSO混合聚类算法的输入数据，这样不仅减少了输入变量的维数，减少聚类的规模，而且保留了原始变量的主要信息，消除变量之间的多重共线性，为具有海量性、高维度性、异构性等特点的

Approach of Web clustering based on hybrid particle swarm optimization model

This paper analyzed the status of Web mining in e-commerce environment, considered the massive and high-dimensional Web data, in order to extract the implicit and unknown knowledge, brought the complexity and the curse of dimensiona-lity. Based on the K-means clustering, particle swarm optimization(PSO) clustering and hybrid PSO clustering algorithms, presented a combination model based on principal component analysis(PCA) and hybrid PSO to cluster log files in the Web servers. The interrelated Web data have been processed by principal component analysis, the results of PCA are input data for hybrid PSO clustering algorithms. It not only reduces the number of input variable and the size of clustering, but also reserve the main information of original variables and eliminates of multicollinearity between the variables; presented an effective mo-del of Web data clustering which have characters of massive, high-dimensional and heterogeneous.