Exact and Hyper-heuristic Methods for Solving the Conference Scheduling Problem

Academic conferences offer notable benefits to participants and promote knowledge advancement. To maximise these benefits, an efficient schedule is crucial. Due to diverse constraints and objectives, various mathematical models and heuristic methods have been developed to meet the unique needs of different conferences. This paper investigates different decision-making tools for creating a generic conference scheduler applicable to a wide range of conferences. We present extended formulations of previously proposed mathematical models to handle constraints at the time slot level, and introduce an approximation model, which is a relaxed version of the previously proposed exact model, obtained through transformations. Additionally, we compare the performance of three methods: a relaxed integer programming model, a decomposed two-phase matheuristic solution approach, and a selection hyper-heuristic algorithm on 16 instances. We highlight the benefits and limitations of each method, providing comprehensive computational results.