An Improved Genetic Algorithm for Team Formation Problem

Abstract

The team formation process is critical to both team performance and team success. And, the key is person-job matching and team collaboration. Successful team formation not only requires team members to meet job requirements, but also systematically consider collaboration among them. However, traditional human resource management relies on qualitative methods is difficult to achieve that. In this paper, we propose a heuristic person-job matching algorithm (HPJMA) and an improved genetic algorithm (IGA) to solve the team formation problem (TFP). First, the heuristic person-job matching algorithm determines whether the job seeker meets all the job requirements. The key idea is to put the right applicants on the right positions. Then, the improved genetic algorithm adopts real number coding to form individuals in the population. Moreover, we utilize a heuristic method to obtain the initial population and then use the elite individual retention strategy to speed up the algorithm convergence. In addition, we introduce the population perturbation strategy to avoid getting struck in the local optimal solution. We aim to maximize the matching degree of the team as a whole. The experimental results show that the proposed algorithms reach a higher overall person-job matching degree on different scales, compared to the three baseline algorithms. Therefore, the proposed heuristic person-job matching algorithm and the improved genetic algorithm can effectively solve the team formation problem.