KNOWLEDGE REPRESENTATION ON STUDENT’S ACADEMIC PERFORMANCE USING RULES BASED METHOD

Abstract

There are various models introduced in recent years that investigated students’ academic performance based on their results, using artificial neural network, rule-based, machine learning, among others. In common scenario at tertiary level education, at least four courses are compulsory for completion by students in the first two semesters, because it would be harder for them to get distinction in academic results when they embark second year onwards. In addition, they need to fulfil course requirements individually or in group, such as tests, quizzes and assignments, ending with exams. It is important to ensure that students are consistently performing throughout their study in ensuring that they graduate on time. With this reason, we propose rule-based method to predict students’ potential success in academic, based on their current progress in coursework. A case study on prediction of students’ performance in assignments is presented by using GUSC factors adapted from personal knowledge management concept, and techniques of decision tree classifier. Analysis is done on a dataset of group coursework results, categorized into Get, Understand, Share and Connect components (i.e. GUSC), which are considered as attributes for classification in performance prediction. The result includes the analysis process of producing frequency patterns by decision tree with information gain classifier.