Abstract
M.Com. (IT Management)
South African Universities are populated by students from different social-cultural backgrounds, countries and high schools. This suggests that these students have different experiences and levels of understanding. It is thus imperative for South African Universities to recognize this diversity in order for them to provide students with better support and equip them with the necessary knowledge so that they will be able to discover their full potential and enhance their skills. Contemporary universities are currently using different practices in order to contribute to the academic performance of students. Some of those practices include employing more tutors, the use of mobile devices for first-year students, employing student assistants and feedbacks measures. These practices are not efficient enough since the number of students who are quitting university is still high. Therefore, universities should embrace other methods that will allow them to improve the academic performance of their students. Data Mining is concerned with generating procedures and studying educational content to facilitate a better understanding of the performance of students. Data mining and artificial intelligence’s goal is to draw knowledge from data through the use of machine learning algorithms. Higher education institutions should embrace the use of Data Mining in order to measure hidden patterns that are affecting the academic performance of students and prevent a low graduation rate. However, in order for universities to use data mining or artificial intelligence, they should first identify the factors that are affecting the academic performance of students and then embed them in the tool that will be used. Algorithms such Bayesian Networks, support vector machines and decision trees can be used to construct a predictive model for the academic performance of students. The best model presented in this study was constructed through the use of support vector machine algorithm. The model does predict the performance of students well in advance of the year-end examinations outcome. Most comprehensive universities may benefit from the model. This study focused on developing a model to predict student performance using quantitative data.