A Framework for Predicting the Students at Risk Using AI: A Case Study

This paper proposes a framework for predicting students who are at the risk of poor academic performance and supports slow learners in higher education institutions (HEIs). Despite numerous studies identifying various circumstances that can lead to a student being at risk, disengagement and slow learning are the major factors that contribute to this issue. Most HEIs aim to provide support to these students after they fall into the students-at-risk (SAR) category which ends as a reactive approach. However, this paper aims to take a proactive approach by developing a framework that can monitor student learning and predict their performance well before they fall into the SAR category. To develop this framework, the authors conducted a case study among the undergraduate students offered in a HEI at Oman. These degrees are awarded by the UK University which is classified as level 6 by the Frameworks for Higher Education Qualifications of UK Degree (FHEQ), Quality Assurance Agency (QAA), UK. The module for this study is selected based on the learning experiences of students registered in the undergraduate courses over the last three years, higher number of registrations and higher number of at-risk students. The proposed framework can be embedded into the existing learning management system (LMS) to predict the future progression of students based on their weekly performance in the classroom and periodic completion of in-class activities. This model also promotes active participation in the learning process as an additional benefit.