Extracting Communities in Networks Based on Functional Properties of Nodes

Abstract

We address the problem of extracting the groups of functionally similar nodes from a network. As functional properties of nodes, we focus on hierarchical levels, relative locations and/or roles with respect to the other nodes. For this problem, we propose a novel method for extracting functional communities from a given network. In our experiments using several types of synthetic and real networks, we evaluate the characteristics of functional communities extracted by our proposed method. From our experimental results, we confirmed that our method can extract functional communities, each of which consists of nodes with functionally similar properties, and these communities are substantially different from those obtained by the Newman clustering method.