欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
研究重叠社区发现技术,可以将重叠社区转化为非重叠的社区发现。通过分裂结点后,使用已有的非重叠社区发现算法来进行社区划分,然后将分裂后的结点还原为源结点即可发现重叠社区。而结点是否分裂由分裂系数来衡量。使用局部的方法来计算结点的分裂系数,并分裂具有较大分裂系数的结点。实验表明此方法可以有效地发现重叠社区,并从时间复杂度上优于基于全局的方法。
英文摘要：
This paper studies overlapping community discovery technology,which can convert any overlapping community detection algorithm into a disjoint community detection algorithm.After getting through the splitting nodes,it divides the community using existing disjoint detection algorithm and restores the split nodes back to the source notes,thus the overlapping community will be detected.The split coefficient can measure whether the node should be split or not.A local approach is employed to calculate node＇s split coefficient,and to split the node with larger split coefficient.Experiments show that this method can effectively detect overlapping community,and is superior to the method based on global optimal in time complexity.
同期刊论文项目
同项目期刊论文
期刊信息
位置：