Atribuição automática de estudantes universitários a turmas baseada em otimização multicritério

Abstract

At FEUP, as in any academic institution, there's a need to section students between different teaching groups according to the capacities of the rooms where practical classes take place. Before the start of each school year, the number of groups and their class schedules are set. Then, students must enroll in them in order to define their individual weekly schedule. This process poses several challenges, however, since students have different preferences regarding which times they want to go to class at and which professors they are assigned to. The current process is divided in multiple phases, the first of which has students being prioritized according to their academic performance and then assigned to the groups they want. Nonetheless, at the end there's still a considerable percentage of unassigned students, which leads into a second phase-which is considered to be quite problematic-where students are assigned on a first come, first served basis. Finally, there's still a third phase in which students must go to the administration office and manually resolve any problems that remain. In this dissertation, we analyze in greater detail the group assignment process used at FEUP, as well as the problems that it brings to its participants. Based on that analysis, we detail the new solution which was developed during the creation of this dissertation. It can automatically assign a considerably larger number of students, giving a big importance to-but not relying entirely on-their preferences. This new system works for assigning groups in not only optional but also mandatory courses, which operate in a slightly different way.