Network Clustering

Abstract

Abstract We use the term network clustering for both the clustering of nodes/links in a network and the clustering of a set of networks. For the clustering of a set of networks, we can use traditional methods. The crucial point is the definition of an appropriate dissimilarity measure between networks. Some options are presented. For clustering nodes/links of a given network, its graph structure can be considered using graph theory notions in the definition of the set of feasible clusterings (clustering with relational constraint), and/or using special criterion functions compatible with a desired network structure (blockmodeling).