Background: In today’s world, technically complex systems can be perceived and presented in the form of network structure. Community detection has a very important application in the domain of social network analysis. Community discovery in networks has grabbed the attention of researchers from multi-discipline. Community detection problem has been modeled as an optimization problem. A large number of existing community detection algorithms adopts modularity as the optimizing function. The measure modularity could not detect communities of smaller size in respect of the size of network i.e. it suffers from resolution limit problem.
Method: This paper addresses the problem of the resolution limit posed by modularity as the fitness function for solving the community detection problem using the discrete bat algorithm. In this work, the discrete bat algorithm with modular density as the optimization function is recommended. Further, widely adopted metric modularity is employed for assessing the quality of community structure.
Results: Experiments are conducted on four real-world datasets. For carrying out the box-plot analysis of the proposed algorithm, ten independent runs are executed. The experimental results show that our proposed algorithm yields to high-quality community structure.
Conclusion: The proposed algorithm overcomes the problem of resolution limit posed by modularity as the fitness function. The results are compared with traditional and evolutionary community detection algorithms. The final outcome shows the superiority of discrete bat algorithm with modular density as the optimization function in respect of a number of communities, maximum modularity, and average modularity as well.