A multiobjective discrete cuckoo search algorithm for community detection in dynamic networks

Evolutionary clustering is a popular method for community detection in dynamic networks by introducing the concept of temporal smoothness. Some evolutionary based clustering approaches need an input parameter to control the preference degree of snapshot and temporal cost. To break the limitation of parameter selection and increase accuracy of detecting communities, we propose a multiobjective discrete cuckoo search algorithm to discover communities in dynamic networks. Firstly, ordered neighbor list method is used to encode the location of nest for population initialization. Secondly, a discrete framework of cuckoo search algorithm is proposed with a modified nest location updating strategy and abandon operator. Finally, based on the proposed discrete framework, a multiobjective discrete cuckoo search algorithm is proposed by integrating the non-dominated sorting method and the crowding distance method. Experimental results on synthetic and real networks demonstrate that the proposed algorithm is effective and has higher accuracy than other compared algorithms.