Applying a mixed objective model in a university timetabling solution searching technique

Course timetable is a problem that can be solved using a search technique. The task is usually required in all educational institutes. Even though several literature´s had been proposed for the problem, the practical solution is still required a lot of human modifications due to a list of specific requirements and the different characteristics of each institution. This work aims to develop a flexible model to guide the search technique namely dept-bounded discrepancy search to solve timetabling problems. The experiments are done using the real data set of the Faculty of Engineering at Prince of Songkla University. The results show that the objective model can guide the search to a solution that satisfies all hard constraints with a minimal number of soft-constraints violations. However, the order of the soft constraints considered by the search technique can affect the number of soft-constraints violations.