Manuscript received January 29, 2024; revised May 14, 2024; accepted July 8, 2024; published September 14, 2024
Abstract—At the Institute of Digital Arts and Sciences at Suranaree University of Technology, students face the crucial decision of choosing between digital technology and digital communication as their professional field. This choice significantly influences their academic pursuits and future career paths. This research aimed to construct and compare the performances of various models in predicting students’ selection of professional fields, utilizing data from student questionnaires at the Institute of Digital Arts and Science. Classification techniques, considered a subset of data mining methods, were applied, and models were constructed using five algorithms: Decision Tree, Naïve Bayes, One Rule (OneR), Support Vector Machine, and K-Nearest Neighbors. These models were evaluated based on accuracy, recall, precision, and F-measure and cross-validated with 10, 20, and 30-fold evaluations. The findings revealed that the Naïve Bayes algorithm-based model, especially with 20-fold cross-validation, was most accurate, achieving 89.6%. The Support Vector Machine algorithm-based model exhibited the highest precision at 82.1% with 30-fold cross-validation. The Decision Tree algorithm-based model achieved the highest recall and F-measure at 83.3% and 81.5%, respectively, with 10-fold cross-validation.
Keywords—professional fields, major selection, prediction, educational data mining, classification
Cite: Phichayasini Kitwatthanathawon, "Predicting the Professional Field of Students Using Data Mining: A Case Study of an Autonomous University in Thailand," International Journal of Information and Education Technology vol. 14, no. 9, pp. 1277-1284, 2024.