Authors :
Jecca Claire Pepe; Shane Bonghanoy; John Christian Azores; Ralph Renson Enisimo; Emmie Faye Marione Matabile; Jeffrey Calim; Johani Basaula
Volume/Issue :
Volume 11 - 2026, Issue 3 - March
Google Scholar :
https://tinyurl.com/ydpunstn
Scribd :
https://tinyurl.com/ptvxkhyu
DOI :
https://doi.org/10.38124/ijisrt/26mar1928
Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.
Abstract :
Student dropout remains a major problem in universities in the Philippines. Although student data is available,
many institutions lack automated systems for early identification of at-risk students. This study developed a Dropout
Decision Support System using a threshold-based rule system combined with the Support Vector Machine algorithm. The
web-based system classifies students into High, Medium, and Low risk categories. Data were collected from the Registrar
and Guidance offices, covering several academic years. Key variables influencing dropout risk included grade point
average, attendance, and financial balance. System evaluation followed a recognized software quality model and involved
all our respondents. Results indicated strengths in usability, functionality, and reliability, showing that the system was
accurate, intuitive, and reliable. Analysis of Variance results, no significant differences among evaluator groups. Model
validation demonstrated strong performance across all risk categories, confirming the system’s effectiveness as an early
warning tool.
Keywords :
Dropout Prediction, Machine Learning, Support Vector Machine, Decision Support System.
References :
- M. Y. Amare and S. Simonova, “Global challenges of students dropout: A prediction model development using machine learning algorithms on higher education datasets,” SHS Web of Conferences, vol. 129, p. 09001, 2021.
- C. Dann, P. Redmond, M. Fanshawe, A. Brown, S. Getenet, T. Shaik, X. Tao, L. Galligan, and Y. Li, “Making sense of student feedback and engagement using artificial intelligence,” Australasian Journal of Educational Technology, 2024.
- I. E. Livieris, T. A. Mikropoulos, and P. Pintelas, “A decision support system for predicting students’ performance,” 2016.
- M. R. Alzahrani, “Predicting student performance using ensemble models and learning analytics techniques,” 2024.
- J. Guzman, “Stakeholders’ participation in School Improvement Plan and school performance of secondary schools,” International Journal of Arts, Sciences and Education, 2022.
- S. Kim, E. Choi, Y.-K. Jun, and S. Lee, “Student dropout prediction for university with high precision and recall,” Applied Sciences, vol. 13, no. 10, 2023.
- J. A. C. Montes, M. C. B. Rodríguez, and M. A. Chamorro, “Drop-out prediction in higher education using imbalanced multiclass dataset,” Journal for ReAttach Therapy and Developmental Diversities, 2023.
- S. S. Padmannavar and S. Begum, “Prediction of student performance using genetically optimized feature selection with multiclass classification,” IJETT, 2022.
- K. Sagar and A. Saha, “Exploring the effect of task difficulty on usability scores of academic websites computed using SUS,” 2020
Student dropout remains a major problem in universities in the Philippines. Although student data is available,
many institutions lack automated systems for early identification of at-risk students. This study developed a Dropout
Decision Support System using a threshold-based rule system combined with the Support Vector Machine algorithm. The
web-based system classifies students into High, Medium, and Low risk categories. Data were collected from the Registrar
and Guidance offices, covering several academic years. Key variables influencing dropout risk included grade point
average, attendance, and financial balance. System evaluation followed a recognized software quality model and involved
all our respondents. Results indicated strengths in usability, functionality, and reliability, showing that the system was
accurate, intuitive, and reliable. Analysis of Variance results, no significant differences among evaluator groups. Model
validation demonstrated strong performance across all risk categories, confirming the system’s effectiveness as an early
warning tool.
Keywords :
Dropout Prediction, Machine Learning, Support Vector Machine, Decision Support System.
