Personalized Elective Course Recommendation Using Content-Based Filtering

Recommending suitable elective courses to students is crucial in order to improve their academic journey and align their learning objectives with their desired careers. Besides, every student is very concerned about selecting appropriate elective courses since that directly impacts their performance in academic studies and especially their career growth. Proposing a system for course recommendations to facilitate students choosing elective courses is therefore crucial. Research on recommender systems is still going strong and is applied in many different fields, yet the recommendation system of course is regarded as a difficult field that has not been extensively looked into. Based on each student's academic background, this research uses content-based filtering approaches to recommend personalized elective courses. This research illustrates a recommendation system that models both elective courses and student profiles as binary vectors, using cosine similarity to analyze the similarity between completed prerequisite courses and available electives to determine the most suitable course recommendations for students. Besides, based on higher similarity, the system recommends top N personalized elective courses. It uses a real-world dataset. Two popular evaluation metrics are employed to evaluate the system's performance: root mean square error (RMSE) and mean absolute error (MAE). The result demonstrates the potential of content-based filtering to recommend personalized elective courses accurately. It thereby enhances student satisfaction, maximizes academic performance, and accelerates course selection.