摘要
随着复杂社交网络的深入发展,有关社交网络社区发现算法的研究也在不断推陈出新,目前有关社区发现算法的研究大多只利用到网络中单个维度的信息。文章提出一种综合考虑用户交互行为和相似度的社区发现方法,通过有效融合多维信息,在社交网络中探测出社区结构。该方法将网络中用户之间的多维关联概括为交互行为和相似度,使用加入相似性惩罚因子的相似模块度作为目标函数来指导社区的划分。在真实数据集上的实验结果表明,该方法不仅能够体现网络中的动态变化,而且还能得到联系紧密、属性相似的节点集合,证明了方法的合理性和有效性。
With the development of ihe complex social networks, researches on algorithms about social networks community detection also develop constantly. Researches on algorithms about social networks community detection only take advantage of single dimension information of the network. This paper presents a community detection method that considers user interaction and the similarity comprehensively, detecting community structure in social network by mixing together multiple dimensions information effectively. The method summarizes multi-dimensional relations between users as interaction and similarity .using similarity modularity that is added similarity penalty factor as object function to guide the community division. Experimental results on real data sets show that the method not only can reflect the dynamic changes in the network, but also can get closely linked collection of nodes with similar attributes, proving the rationality and effectiveness of the method.
出处
《信息网络安全》
2015年第7期77-83,共7页
Netinfo Security
基金
国家自然科学基金[61402112]
福建省安全课题[828398]
关键词
社交网络
社区发现
相似模块度
social networks
community detection
similarity modularity