基于社团密合度的复杂网络社团发现算法

传统的社团发现算法大多存在划分效果和复杂度相矛盾的问题，为了解决该问题，提出一种新的单社团结构评价标准——社团密合度(group density).在此基础上，设计了一种基于凝聚思想的社团发现算法，该算法通过不断融合小社团，使网络的社团结构向平均社团密合度最大的方向发展，并使用模块度检测算法的划分结果.通过与经典的GN，Fast Newman，LPA等算法对多个数据集进行实验对比，验证了本文算法在获得较好的划分效果的同时具有较低的时间复杂度.

Community Detection Algorithm for Complex Networks Based on Group Density

Most of the traditional community detection algorithms cannot balance partitioning effect and complexity well. So， this paper presents a new evaluation standard of single community called group density. Based on the group density， a community detection algorithm based on agglomeration is proposed. The algorithm continues to integrate small communities， and makes the community structure of the network develop in the direction of maximizing average group density. Modularity is employed to detect the partitioning effect of the algorithm. Experimental results demonstrate that the new algorithm outperforms the traditional GN， Fast Newman， LPA algorithms in multiple data sets， which shows that the algorithm proposed has better partitioning effect and lower time complexity.