Clustering sets of objects using concepts-objects bipartite graphs
SUM'12 Proceedings of the 6th international conference on Scalable Uncertainty Management
| Hi-index | 0.00 |
Communities in unipartite networks are often understood as groups of nodes within which links are dense but between which links are sparse. Such communities are not suited to bipartite networks, as there is only one-to-one correspondence between communities of different types. Recently, B. Long et al. Introduced the link-pattern based community, which allows many-to-many correspondence between communities. In this paper, we propose a measure for evaluating the goodness of different partitions of a bipartite network into link-pattern based communities. Such a measure is useful for both comparing various community detection methods and devising new community detection algorithm based on optimization. We demonstrate the effectiveness of the proposed measure using the famous Southern women bipartite network.