Algorithmic Analytics for Outcomes-based Tertiary Education Performance Assessment

Abstract

Global educational institutions are continuously working to enhance the needed curricula to answer the demands of state and business enterprise. They are now working with Outcomes-based Education (OBE), which denies the traditional way of education to students. Instead, it makes students show that they are capable of doing the required learning outcomes. The quest for learning the correct assessment for education while under the era of OBE is still an on-going call. With the help of Educational Data Mining, a real-world dataset of 549 were trained from the student record. The study intends to identify the probability of the student to pass practical evaluation following OBE methodology, using selected WEKA-based classifiers, namely Naive Bayes, C4.5, and Random Forest. The said classifiers revealed promising accuracy of 78.10% for Naïve Bayes, 93.07% for C4.5, and 95.99% for Random Forest. Hence, the EDM's future is really for further study, which can be implemented in the fields of entertainment, industry, medicine, and many others because of massive raw databases.