Decision-making strategies for multi-objective community detection in complex networks

Abstract

Multi-objective optimization algorithms have demonstrated their effectiveness and efficiency in detecting community structure in complex networks, by which a set of trade-off partitions of networks are obtained instead of a single partition. The large number of partitions lead to a challenging problem for decision makers: how to obtain an ideal partition from the set of trade-off partitions of networks. In this paper, we present two decision-making strategies for obtaining the ideal partition, one is based on the knee points and the other is based on the majority voting. Experimental results on random networks and real-world networks illustrate that the presented two strategies are very competitive for obtaining an ideal partition of networks.