Optimizing the University Student Advising Process through Weighted Minimum Spanning Tree Algorithm

  • Danilo Cabildo Sabado College of Computing and Information Sciences, University of Makati, Philippines


Purpose – This paper discusses a new approach combining the network graph theory on minimum spanning tree and ensemble-based machine learning algorithm for weight generation to solve an optimization problem for common university advising processes. University advising requires an expert’s validation to identify an optimal set of courses that can be recommended to students to finish their program of study at an earliest time while giving heavy considerations on the volume of workload for a given semester.

Method – The study was analyzed utilizing the historical dataset from students’ academic profiles from a local university from the Philippines. The process undergoes a set of weight generation, MST model application and ensemble scoring method.

Results – The proposed algorithm was validated using the expert validated advising results which yielded to 92% accuracy.

Conclusion – This paper attempts to develop a solution to the tedious process of student advising through representing the curriculum in a directed graph and implementing a minimum spanning tree algorithm with ML-based weight edges.

Recommendations – To address the potential reasons on the detected accuracies, it is further recommended to include all other constraints such as accommodating co-requisites between courses.

Practical Implications – The proposed algorithm can be implemented in a web-based application or integrated in an enrollment information system for a more justifiable appreciation of the envisioned process improvement.

How to Cite
SABADO, Danilo Cabildo. Optimizing the University Student Advising Process through Weighted Minimum Spanning Tree Algorithm. International Journal of Computing Sciences Research, [S.l.], v. 7, p. 1723-1736, feb. 2023. ISSN 2546-115X. Available at: <//stepacademic.net/ijcsr/article/view/370>. Date accessed: 28 sep. 2023.