Modification of University Examination Scheduling Model: Elevating Timetabling Communities’ Satisfaction

Abstract

The examination scheduling problem involves assigning exams, invigilators, and students to specific venues and time slots. It is a challenging task for all institutions even in its original form, especially when handling a huge number of examinations. The complexity increases when it requires adhering to various requirements. Many attempts have been made to solve the problem. However, they do not encounter user satisfaction. Models are constructed primarily based on the basic requirements of the respective university, and an attempt has been made to address these matters. Nevertheless, it took a considerable amount of computational time to reach 87.61% optimality. Factors causing the long computation time were identified, and an improved model was developed, emphasizing user considerations by including requirements for the full utilization of the main venues. In the new model, all possible causes were addressed. Exam assignments were optimized to match resources, schedule more exams in earlier time slots, and indirectly maximize venue usage. The objective function is reformulated to reduce the computational time while improving the optimality gap. The model was constructed using Mixed-Integer Linear Programming (MILP) and solved with the Advanced Interactive Multidimensional Modelling System (AIMMS) software, utilizing the CPLEX 20.1 solver on a Core i7 computer (3.40 GHz, 16GB RAM). Validation using a real examination problem, achieved 98.1% optimality, while further enhancing user satisfaction. A detailed analysis and comparison are discussed. This mathematical model will benefit most educational institutions by satisfying the majority of community preferences in producing better examination schedules in the future.