Discovering Academic Pathways of Course Enrollment with Unsupervised Clustering and Rule Mining
Students & Supervisors
Student Authors
Supervisors
Abstract
Student course enrollment data often contains hidden patterns that can provide insights into academic behavior and curricular structures. However, detecting meaningful associations and clusters from such data remains challenging due to the overlap of courses across semesters and specializations. This study presents a combined framework of association rule mining and clustering for analyzing student enrollment records. A binary transaction matrix was created from the most frequent courses and analyzed using both pairwise co-occurrence measures and Apriori rules to capture complementary associations. For segmentation, K-Means, Mean Shift, and Gaussian Mixture Models were applied, with evaluation through silhouette score, Davies Bouldin index, and Dunn index. Results showed that Mean Shift achieved the highest cluster separation but produced too many clusters, while Gaussian Mixture provided fewer interpretable clusters aligned with curricular tracks. Pairwise methods identified strong but rare associations, while Apriori captured more frequent rules. Together these findings demonstrate that student enrollments are not random, but form organized trajectories based on semester timing and subject specialization. The framework supports curriculum design, targeted advising, and identification of academic pathways, providing a practical tool for institutions to enhance program management and student guidance.
Keywords
Publication Details
- Type of Publication:
- Conference Name: International Conference on Computer and Information Technology (ICCIT 2025)
- Date of Conference: 19/12/2025 - 19/12/2025
- Venue: Long Beach Hotel, Cox’s Bazar, Bangladesh
- Organizer: IEEE