社交网络的迅猛发展极大地方便了人们的日常生活、工作和学习,但也带来了大量复杂的交互行为和连接模式.如何有效地综合分析网络中的交互信息和网络节点之间存在的连接信息,进而完成高效的社团检测,是在当前网络多维属性的复杂背景下进行网络分析所面临的关键难题.基于此,从有效融合两类不同的异质信息研究出发,提出了一种基于交互行为和连接分析的社交网络社团检测(CDUILS)方法.该方法基于两类信息能够从不同的角度反映网络同一个社团归属的假设,采用联合非负矩阵分解架构,以迭代更新的方式,同时利用两类信息进行社团结果的获取.在真实网络数据集上的实验表明,与已有方法相比,所提方法能够有效融合两类信息进行社团检测,取得了更好的社团划分质量.%With the rapid development of social media network,the user is also more convenient to participate in social networking,which also brings a large number of complex interaction and connection mode.How to effectively analysis the interactive information and the connection information between network nodes to complete the efficient community detection is the key problem faced by current network analysis.Based on this,this paper put forward a kind of social network community detection method(CDUILS) based on the interaction behavior and link analysis.In this method,the interaction information between nodes is used as the cooperative learning of the community.The non negative matrix factorization is used to analyze the two types of information sources by the way of iterative update,and the community results can be obtained with two kinds of information retrieval.Experiments on real data sets show that the proposed method can effectively utilize the interaction behavior to guide the community division and have better quality of community division.
展开▼
