摘要
为了能够更准确地对邻域重叠网络进行社团结构探测,研究人员对基于完全子图的社团探测算法进行了改进。在合并完全子图团簇时,计算每一对完全子图的重叠节点个数,设置合并完全子图的阈值,如果大于阈值,则合并。在处理不在团簇内的其他节点时,采用按照比例系数大小来划分规则进行划分。算法应用于空手道俱乐部和科学家合作网当中,验证算法可以更准确地探测邻域重叠社团结构。
For detecting the community structure of neighbourhood overlapping network more precisely, the community detection algorithm based on complete subgraph was improved. When merging the complete subgraphs, the quantity of over- lapping nodes between each pair of subgraphs is calculated and the threshold value is set. The complete subgraphs are merged if the amount of overlapping nodes is larger than the threshold value. The rule of partitioning the scale factor is used to divide the other nodes which are not in any cluster. The algorithm is used in Zachary's karate club and scientific collaboration net- works. The actual application proves that the algorithm can detect the overlapping community structure more accurately.
出处
《现代电子技术》
2012年第18期114-118,共5页
Modern Electronics Technique
关键词
邻域重叠网络
完全子图
社团结构探测
比例系数
neighbourhood overlapping network
complete subgraph
community structure detection
scale factor