Community structure detection in networks based on Tabu search

Abstract

Community structure is a main structural feature of networks, which can model many complex systems. It illustrates a lot of information about networks like their internal organisation and the degree of similarity between network nodes. Many methods have been proposed to find community structure in networks. However, there is always requirement to new methods for many reasons especially the accuracy. In this paper, we present a new method to find community structure in networks. Our method is hierarchical algorithm based on Tabu Search metaheuristic. A dendrogram is built by our method after dividing a network several times. The community structure will be selected from the built dendrogram based on quality function called Modularity. Our method has been tested on artificial benchmark and real networks. Results demonstrate the superiority of our method compared with several state-of-art methods.

Keywords:

• Community detection
• networks
• Tabu search
• normalised mutual information
• modularity

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Bilal Saoud

Bilal Saoud holds a Ph.D. degree in Computer Science from the University of Bejaia, Algeria (2017), Magister degree (post- graduation) in Computer Science from the University of Bejaia, Algeria (2012). He was a lecturer at the University of M'sila, Algeria (2012–2013). He was a military officer during his military service. He is a teacher–researcher at the University of Bouira, Algeria. His research areas include Ad Hoc network, wireless sensor network, social network and community detection in network, statistical models and copula, computational intelligence.