The application of big data analytics to improve students’ performance in South Africa

Abstract

The thesis proposes a normative model for predicting students’ performance at a selected tertiary Institution utilizing big data analytics (BDA). BDA can transform education by uncovering patterns, correlations, and other insights into learning outcomes, students’ performances, and the effectiveness of learning and teaching practices. Therefore, the aim of this research was to explore how BDA and Machine Learning can be used within South Africa Higher Education and Training to improve students’ performance and the challenges thereof. Adaptive structuration theory (AST) was used as a lens on the problem to first understand and interpret the embedded socio-technical processes. The Cross-Industry Standard Process for Data Mining (CRISP-DM) Model was used to devise a plan for the implementation of the machine learning algorithm to meet the objective of predicting students’ performances. The algorithm was applied to raw data gathered from different cohorts of students on a structured postgraduate qualification. The implication of the output suggests the establishment of a physical Information Technology infrastructure that makes use of DBA capable of pulling information (structured and unstructured) into a presentation layer that allows for the application of ML on relevant students’ data.