基于聚类分析的SNS网络研究

Study on SNS Network Based on Cluster Analysis

SNS网站积累着起巨大的用户量以及用户的关系网络,蕴藏着巨大的商业价值。以SNS网站社会网络为研究对象,结合聚类分析和社会网络分析的基本思想,提出了以社会心理学理论为基础的针对SNS关系分析的框架——基于聚类分析的社会网络分析框架SNAC,定义了SNS关系权重指标"亲密度",基于"簇饱和度"改进K-Means聚类算法,用网络参数描述了SNS关系网络的特征和小世界特征,利用实验验证了分析框架的可行性。

SNS Sites accumulate a huge number of users and user's relationships,containing tremendous business value.Using the basic idea of cluster analysis and social network analysis,this paper studies social network of SNS sites,and puts forward a analytical framework of SNS relationships analysis,SNAC,based on the modern rumor psychology theory.SNAC defines the intimacy indicator to weight relationship between SNS users,improves the K-Means clustering algorithm based on member saturation,and describes the characteristics and small-world network characteristics of social network of SNS sites.An experiment was conducted to verify the feasibility of analytical framework.