Optimizing Fair and Efficient Group Formation in Community Service Program Using Particle Swarm Optimization

Abstract

This study aims to evaluate the performance of the Particle Swarm Optimization (PSO) algorithm in solving the problem of group formation for community service program (Kuliah Kerja Nyata). The motivation for this research stems from the need for an efficient and effective method to optimally form KKN groups based on specific criteria. The research problem is how PSO performs in terms of computational time, convergence speed, and the quality of solutions produced. The method employed involves testing the PSO algorithm on 27 groups with 10 independent runs, measuring computational time, optimal iterations, initial scores, and best final scores. The results show that PSO is highly computationally efficient, achieving exceptionally fast average execution times and demonstrating remarkable convergence, with most groups reaching optimal solutions in the first iteration. However, two distinct patterns emerged: several groups achieved low optimal scores (95–97), while most other groups (16–26) stagnated at extremely high initial scores (100000.0) without improvement. These findings suggest that although PSO is a fast and effective method, its performance is highly dependent on problem characteristics and initialization. As a recommendation, further studies are needed to identify the factors causing stagnation at high scores and to test parameter adjustment or alternative initialization strategies to enhance the algorithm’s ability to escape local minimal. The contribution of this research is to provide a deeper understanding of the potential and limitations of PSO in the context of group formation, which may serve as a foundation for developing more robust optimization algorithms in the future.