الاكتشاف التلقائي لمعادلة مقررات الطلاب باستخدام التعلم الآلي

Abstract

Comparing the educational curricula of different courses across various universities and educational institutes is a complicated and difficult task. However, technological and digital tools can help develop new and effective methods to equate educational courses. Hence, the researcher of this study aimed to discuss the models that have been implemented in this context in the past and developed a new model with greater effectiveness than the previous ones. The developed model addresses the problems within traditional course equivalence methods by using a supervised machine learning (SML) algorithms. The study used a specific dataset to compare courses at Najran University and Umm Al-Qura University. The dataset contains 965 rows and 18 columns and applied SML algorithms in Orange tool. The results were measured based on several criteria, including area under ROC curve (AUC), classification accuracy (CA), F1 measure, precision, and recall. The results showed that based on AUC, the best algorithms are support vector machines (SVM), while based on CA they are SVM, K- nearest neighbors (KNN), and random forest. Additionally, based on precision, recall, and F1 measure the best algorithms are SVM, KNN, random forest, and logistic regression.