A Particle Swarm Optimization-Based Heuristic for Software Module Clustering Problem

The large-scale software module clustering problems (SMCPs) are very difficult to solve by using traditional analytical/deterministic-based optimization methods due to their high complexity and computation cost. Recently, particle swarm optimization (PSO) algorithm, a non-deterministic meta-heuristic search algorithm, gained wide attention and has been adapted to address the various large-scale science and engineering optimization problems. However, the applicability and usefulness of PSO algorithm have not been studied by any researcher till date to solve the SMCPs. In this paper, we introduce PSO-based module clustering (PSOMC), which partitions software system by optimizing: (1) intracluster dependency, (2) intercluster dependency, (3) a number of clusters, and (4) a number of module per cluster. To this contribution, we redefine the terms “position” and “velocity” of original PSO under the discrete scenario that best suited to SMCPs. To demonstrate the performance of the proposed approach, extensive experiments on six real-world SMCPs are carried out. We also compare our approach with existing state-of-the-art software module clustering meta-heuristic approaches (group genetic algorithm, hill climbing, and simulated annealing algorithm). The experimental results show that the proposed approach is effective and promising for solving SMCPs.