A Clustering Coefficient to Identify Important Nodes in Bipartite Networks.

Bipartite networks have gained an increasing amount of attention over the past few years. Network measures in particular, have been the focus of this research as many of them cannot be directly applied to bipartite networks. The clustering coefficient is one measure that has been redefined recently to suit the analysis of bipartite networks. Building up on this definition, we propose a clustering coefficient that distinguishes between differently structured bipartite clusters. We use this measure to identify influential nodes in a given bipartite network. By comparing the global and local clustering coefficients, we assign a score to each node that indicates the extent to which it drives the clustering behaviour of the whole network. We demonstrate that our clustering coefficient is not only able to identify influential nodes, but gives new insights into a network’s structure.