CAJ | 학술논문

大数据环境下社交网络的社团结构研究对解决很多现实问题有着重要的意义。社团通常被看作是有相对紧密的内部连接和比较稀疏的外部连接的子图，重叠是社团结构的一个重要特征。论文基于 G（n ，p）模型，提出了一种生成包含重叠社团的合成网络的方法，然后基于 k-均值和随机游走设计了一种重叠社团的检测算法，并在合成网络上初步验证了该方法的可行性。
대수거배경하사교망락적사단결구연구대해결흔다현실문제유착중요적의의。사단통상피간작시유상대긴밀적내부련접화비교희소적외부련접적자도，중첩시사단결구적일개중요특정。논문기우 G（n ，p）모형，제출료일충생성포함중첩사단적합성망락적방법，연후기우 k-균치화수궤유주설계료일충중첩사단적검측산법，병재합성망락상초보험증료해방법적가행성。
Community structure in social networks is of great significance to many real world problems under the big data environment .Communities of a social network are regarded as connected components with comparatively dense internal connections and sparse external connections .Overlapping is a key feature of the community structure .Based on the G(n ,p) model ,a method is presented to generate synthetic networks having overlapping communities ,and an overlapping community detection algorithm basing on k-means and random walk is proposed .Experiments on synthetic networks show the efficiency of the proposed algorithm .