Teaching Schedule Optimization Using Tabu Search Algorithm in University Academic Timetabling
Аннотация
Timetabling processes at universities are complex combinatorial problems that are highly constrained and often involve optimizing fair and efficient schedules as part of the timetabling process. The timetabling process involving academic schedules focuses on allocating instructors, courses, and rooms, which should be conflict-free and allocate resources fairly for the students (the courses), instructors, and university (the rooms). Most existing academic timetabling methods struggle to deal with scalability, flexibility in constraints, and convergence to the optimal or near-optimal solution. These limitations tend to exacerbate scheduling conflicts and, therefore, ultimately lead to low satisfaction levels among various stakeholders. The gaps and limitations of current scheduling methods are addressed in this study through a formal scheduling framework, which utilizes the Tabu Search Algorithm (TSA), unlike traditional methods. TSA is a metaheuristic optimization procedure that utilizes adaptive memory and a tabu list to avoid local minima and to explore the solution space effectively. The proposed system, TSA, creates academic schedules or timetables for students, the community, and instructors while addressing hard constraints like instructor course availability, room capacity, and soft preferences. At each iteration, TSA attempts to modify feasible solutions' timetables while avoiding all previously visited suboptimal configurations. The experimental results reveal that TSA achieves the fewest scheduling conflicts (as low as 2), efficient load balancing (SD as 2.3), high soft constraint satisfaction (88–90%), and acceptable execution time of ~6s, outperforming conventional methods significantly. The TSA-based model represented an effective way to enhance the reliability and flexibility of academic timetabling.