 |
    |
A Cooperative Evolutionary Approach to Learn Communities in Multilayer Networks
Alessia Amelio and Clara Pizzuti
National Research Council of Italy (CNR), Institute for High Performance Computing and Networking (ICAR), Via P. Bucci 41C, 87036, Rende, CS, Italyamelio@icar.cnr.it
pizzuti@icar.cnr.it
Abstract. In real-world complex systems objects are often involved in different kinds of connections, each expressing a different aspect of object activity. Multilayer networks, where each layer represents a type of relationship between a set of nodes, constitute a valid formalism to model such systems. In this paper a new approach based on Genetic Algorithms to detect community structure in multilayer networks is proposed. The method introduces an extension of the modularity concept and adopts a genetic representation of a multilayer network that allows cooperation and co-evolution of individuals, in order to find an optimal division of the network, shared among all the layers. Moreover, the algorithm relies on a label propagation mechanism and a local search strategy to refine the result quality. Experiments show the capability of the approach to obtain accurate community structures.
LNCS 8672, p. 222 ff.
lncs@springer.com
© Springer International Publishing Switzerland 2014