A Metaheuristic-Based Modularity Optimization Algorithm Driven by Edge Directionality for Directed Networks

Community detection of directed networks has been demonstrated to be able to reveal network function and dynamic characteristics, due to the special information brought by asymmetric edge directionality, but the critical issue of improving the quality of detected communities has not been fully addressed. In this study, we propose a metaheuristic-based modularity optimization method for directed network community detection (MMDCD), which comprehensively leverages nature-inspired metaheuristics and edge directionality information. Specifically, a nature-inspired metaheuristic algorithm, termed diversified farmland fertility (DFF), is improved to provide an efficient optimization strategy for directed modularity maximization. More importantly, the information induced by edge directionality, including edge density, edge pattern, and node importance, are fully used to improve the quality of optimized partitions. Experimental results on synthetic and real-life networks prove the effectiveness of MMDCD, which is superior to state-of-the-art community detection algorithms in terms of quality and accuracy.