A general framework for testing different student team formation strategies

Abstract

One of the most important problems faced by teachers is grouping students into proper teams. The task is complex, as many technical and interpersonal factors could affect team dynamics, with no clear indication of which factors may be more relevant. Not only the problem is conceptually complex, but its computational complexity is also exponential, which precludes teachers from optimally applying strategies by hand. The tool presented in this paper aims to cover both gaps: first, it provides a range of grouping strategies for testing, and second, it provides artificial intelligence mechanisms that in practice tone down the computational cost of the problem.