论文部分内容阅读
聚类分析是w eb社区发现的主要方法之一,近年来基于谱图理论的谱聚类由于其诸多优点而备受关注,在介绍谱聚类基本理论的基础上,着重分析了包括SM、N JW、NystormCut与KASP在内的4种典型谱聚类方法的基本流程与优缺点,在聚类有效性与时间效率上对4种算法进行实验比较,实验结论为w eb社区发现工程实践提供了借鉴.
Clustering analysis is one of the main methods for community discovery. In recent years, spectral clustering based on spectral theory has drawn much attention because of its many advantages. Based on the introduction of the basic theory of spectral clustering, N JW, NystormCut and KASP, four kinds of typical spectral clustering methods of the basic processes and advantages and disadvantages of the clustering effectiveness and time efficiency of the four algorithms for experimental comparison, the experimental results for the w eb community discovery engineering practice to provide Learn from.