二分网络社团结构的比较性定义

Comparative Definition of Community in Bipartite Network

为了解决二分网络的聚类问题,借鉴单顶点网络社团结构的比较性定义,提出了直接基于原始二分网络社团结构的比较性定义,这个定义允许社团之间存在重叠,并定义了在二分网络中顶点与社团之间的作用力。在不引进额外参数的情况下,设计了基于此定义的二分网络的聚类算法,并将此算法应用于人工网和一些实际网络中,结果表明这个算法可以比较准确地对网络进行聚类,说明该定义是有效的。

In order to know the community structure in bipartite network,we propose a comparative definition of community in bipartite network, overlapping between communities is allowed in this definition. We also define the attractive force between the node and the community in bipartite network. Then without any extra parameters, a new cluster algorithm of community in bipartite network is proposed,and we run it in the artificial networks and some real networks. The algorithm runs quite well in detecting communities according to the result, so this definition of community is effective.