This paper demonstrates how each of these methods can divide the structure of large-scale network.

Problem Resolved

This paper has described several methods for the analysis of citation networks, which are acyclic directed graphs of citations between documents. Using the network of citations between opinions handed down by the US Supreme Court as an example, this paper described and demonstrated three analysis techniques. The first makes use of a probabilistic mixture model fitted to the observed network structure using an expectation–maximization algorithm. The second is a network clustering method making use of the recently introduced method of modularity maximization. The third is an analysis of the patterns of time variation in eigenvector centrality scores, particularly the “authority” score introduced by Kleinberg.