Recommendation System for Collegian Student's Weekly Course Schedule

Abstract

Existing research on course scheduling was conducted only from the institutional side.  However, students usually have other considerations, such as routine activities outside of class, course time, holidays in a week of study, and lead time between courses. These conditions had never been taken into consideration in existing research.  In this paper, a recommendation system was proposed using Depth First Search and Simple Multi Attribute Ranking Technique methods.  Depth First Search method was used to find all possible alternative schedules. All the possible alternative schedules were used to determine the schedule that best suits student preferences using Simple Multi Attribute Ranking Technique method. The system performance was measured through simulation to get course schedule recommendations for 28 students.  The simulation results were compared with the ideal schedule desired by the students and the real course schedule for those students. The accuracy of the recommended schedule against the ideal schedule desired by students was 70.8% with an average processing time of 1.05 seconds. The accuracy of the recommended schedule increased to about 91% when compared to the actual student courses schedule.  So it can be concluded that the research can help to recommend students' weekly class schedules in real terms.