An Artificial Intelligence Approach to Monitor and Predict Student Academic Performance

Abstract

Recent progress in technology has altered the learning behaviours of students. It can easily be said that the improvements in technologies especially from the area of artificial intelligence and data mining can empower students to learn more efficiently, effectively, and contentedly. The purpose of this paper is aimed at developing a prediction model that gives a guide to the stakeholders likes university, student’s sponsor and parents that can monitor and interpret the on-going student’s performance. The proposed system will help the user to identify students who have a high risk or low risk to ending the semester with unsatisfactory results through discovering the essential features that can influence student’s academic performance. The methodology used were mixed method approach which is qualitative and quantitative. For the part of quantitative technique, the data set for this research will be collected from the academic records section. Meanwhile, the part of qualitative will go through a few main stages which are focused on developmental phase. The prediction model used in this research are RepTree, k-NN and Naïve Bayes. The finding states that RepTree had the highest accuracy compared to k-NN and Naïve Bayes techniques. It also suggests that the RepTree algorithm would be the best prediction model which has focused the research in term of developing the system of monitoring students’ academic performance at UPTM.