机构地区: 福建师范大学软件学院
出 处: 《小型微型计算机系统》 2013年第7期1531-1533,共3页
摘 要: Web文档聚类是web数据挖掘的重要任务之一,针对Web文档向量空间的高维性与数据聚类问题的最优化性质,采用LDA对文档向量空间进行降维,提出运用混合优化算法GA_PSO在此低维空间进行寻优,来发现Web文档集的最优簇结构.通过在真实数据集20Newsgroups的实验,结果表明我们的方法具有良好的聚类有效性,能较完全和准确地将主题相关的Web文档聚成一类. Clustering web documents is an important task of web data mining.For high dimensionality of web document vector space and optimization essence of data clustering,a hybrid optimization algorithm based on GA and PSO for clustering w eb documents is proposed,in w hich w eb documents are represented by vector space model,LDA approach is applied to reduce the document vector space.The experiment results from true dataset 20New sgroups show that the proposed clustering algorithm has good clustering performance and the w eb documents focus on a subject are rather completely and exactly clustering together.
领 域: [自动化与计算机技术] [自动化与计算机技术]