Incorporating Recommended Course Prerequisites into Similarity-Based Course Recommender Systems for Higher Education

Abstract

Course selection in higher education, particularly for elective courses in study programs with high degrees of freedom, can be overwhelming for students. While course recommender systems can support students in the process, these systems often overlook recommended prerequisites specified in the curricula of study programs. To address this, we modeled recommended course prerequisites based on module catalog information and developed a collaborative filtering-based recommender system that incorporates prerequisite fulfillment as a quality. We conducted a user study to evaluate the recommender system’s usability and the perceived plausibility of its generated recommendations, comparing it to a matrix factorization-based approach. Our findings indicate that while matrix factorization is generally preferred, partly due to its reduced susceptibility to popularity bias in the data, students generally accept the integration of recommended prerequisites and their impact on explainability of r ecommendations. These results highlight the potential of prerequisite-aware recommendations in enhancing course selection.