在线社会网络的动态社区发现及其演化

发布时间：2019-03-07 21:56

【摘要】：分析了目前动态社区发现及其演化所存在的问题,提出了一种新的动态社区演化方法.该方法利用静态社区挖掘算法提取不同时间快照的每个社区,然后计算出相邻快照的社区之间的演化影响力,进一步分析连续快照中社区结构的发展演化过程.在新浪微博、网络测量Gnutella等大规模实验数据集上的验证,证明了该方法的有效性.此外,实验中还分析了社会网络中节点的出现和消失的频繁程度会影响社区稳定性以及社区结构的演化.
[Abstract]:This paper analyzes the problems existing in the discovery and evolution of dynamic community, and puts forward a new method of dynamic community evolution. In this method, the static community mining algorithm is used to extract each community with different time snapshots, and then the evolution influence between the communities of adjacent snapshots is calculated, and the evolution process of community structure in continuous snapshots is further analyzed. The validity of this method is proved by experiments on Sina Weibo, Gnutella and other large-scale experimental data sets. In addition, the frequency of the emergence and disappearance of nodes in the social network will affect the stability of the community and the evolution of the community structure.
【作者单位】：中国人民大学信息学院;淮阴师范学院计算机科学与技术学院;
【基金】：国家自然科学基金资助项目(71271211);国家自然科学基金重点资助项目(71271211)
【分类号】：TP393.09

【相似文献】