In the contemporary university environment, students have access to various strategies for engaging in self-directed learning. However, they are not required to acquire all knowledge independently which possibly leads students to lose motivation, comprehension, and learning efficiency. To solve this problem, a solution that paired study partners by utilizing Stable matching theory (SMT) was proposed. SMT was applied to analyze the formation of mutually beneficial study partners and optimize the allocation of resources based on specific characteristics. Moreover, SMT was used to optimize allocations, such as determining members’ abilities and skills to maximize overall benefits and tackle complex relationship problems in groups. Our new approach combined the SMT with Non-dominated Sorting Genetic Algorithm III (NSGA-III), to perform multi-objective matching optimization and balance the efficiency between conflicting objectives. This paper is expected to contribute an innovative algorithm-based approach to match students with compatible study groups, potentially enhancing academic collaboration and learning outcomes.