Improving Equity and Access to Higher Education Using Artificial Intelligence
Saoussen Nour El Imen Cheddadi, Mourad Bouache
Abstract
With data on education becoming more available, it has created a strong link between big data and learning analytics. This link paved the way for researchers in these fields to address several issues within higher education. In this study, we address the inequity in higher education, which is a long standing issue, that is depriving many skilled people of the fundamental right of education due to their background, gender, race, ethnicity or social class. Research has shown that deprived groups in terms of their demographics are more likely to fail. In this regard, we propose to use a deep neural network (DNN) that aims to predict students’ final results. This prior knowledge of students’ outcome could be used in favor of underprivileged groups in order to assist and enhance their chances to be successful. The conceived DNN consists of a binary classifier that is trained on Open University Learning Analytics (OULA) dataset, with the aim to classify students into two categories: fail or pass, given a specific module and based on their demographics and interaction with the virtual learning environment. The obtained results are promising and show that deep learning could be an effective tool to promote equity in higher education. Moreover, they reveal that this latter is also a powerful tool for exploring big data within learning analytics, provided that deep neural networks become more accurate when they are provided with additional data.