A Mathematical Model for a Realistic Course Scheduling at Universities

Abstract

Timetabling is a combinatorial and challenging problem in different fields where it is required to allocate scarce resources. One of the main variants is the University Course Timetabling Problem (UCTP), where courses must take a location and the academic sessions must be distributed avoiding overlapping schedules or preference conflicts. Despite the high complexity of its solution, it is increasingly necessary to include more realistic characteristics and formulate efficient models. In this paper, new realistic features focused on the development of student’s curriculum are included and a mixed-integer linear programming model is proposed to address this problem. The model’s objective is to maximize all teacher’s preferences and reduce the overlaps between courses that can be viewed either in advance or delayed. The proposed model was validated in the timetabling in the Industrial Engineering Program at the Pontificia Universidad Javeriana Cali, Colombia. The results show an increase of 47.7% regarding the preferences of the professors, as well as a reduction of 99.37% of the time associated with the semester scheduling process and currently, the perception of students is being assessed as they progress through the semesters, with preliminary results indicating an increase in satisfaction.