CASS (College of Arts and Sciences Scheduler) A MULTI-OBJECTIVE GENETIC APPROACH ON UNIVERSITY TIMETABLING

Abstract

Timetabling is a problem familiar to any educational institution. In general, most schools still implement the manual timetabling, which aside from exhausting significant resources like time, labor and paper, does not assure an optimal solution. CASScheduler (College of Arts and Sciences Scheduler) implements an intuitive university timetabling system, which makes use of Artificial Intelligence techniques. Since the timetabling is a multi- objective problem, at its core is the Non-dominated Sorting Genetic Algorithm II (NSGA-II), a multi-objective evolutionary algorithm, which aside from being fast and elitist, has ability to discern the fitness of a solution over an assortment of objectives, instead of using a singular fitness function characterized by weights. It is able to rank and generate a set of Pareto-optimal solutions, aside from optimizing a single solution, giving the user the best possible alternatives. The project was developed and tested using real-world data, with the College of Arts and Sciences in the University of the Philippines Manila as the pilot setting, which happens to be the largest college in the university. CASScheduler may aid in improving efficiency in resource allocation and schedule assignment, upon its implementation, and also enhance productivity among the college's personnel.