Artificial intelligence-enabled decision support system for South African higher education institutions

Abstract

This study aimed to develop a prototype artificial intelligence-enabled decision support system tailored for South African higher education institutions, recognising the growing significance of such systems in the dynamic landscape of universities. As universities evolve beyond traditional roles into data driven entities, decision-making has become increasingly complex, requiring real-time insights, connectivity and automation. However, the implementation of AI in higher education remains limited. Consequently, there is an actual need for AI-enabled decision support systems within university ICT departments to enhance operational efficiency and provide timely decision-making. This study sought to address this gap by creating an intelligent system that harnessed university data and adapted to changing circumstances, offering prompt, efficient and high-quality service. By relieving ICT support personnel from routine tasks and minimising downtime, this AI-enabled system aimed to enhance customer satisfaction. The research question for this study was how can an AI-enabled decision support system be developed for decision-making within the ICT department at the university? The problem's relevance lies in the need for efficient decision-making processes in South African higher education institutions, as demonstrated by this study using the ICT department as unit of analysis. By leveraging the Design Science Research methodology, this study integrated Architectural design theory and Decision theory in developing the artefact. Ontological pragmatism and intersubjective epistemology were employed to address an existing real-world problem. The study initiated semi-structured interviews to identify challenges within the university's ICT department. Subsequently, the AIDSS prototype was developed. This prototype incorporated business automation, preventive asset maintenance, and predictive analytics functionalities to comprehensively address the identified issues. Business automation aimed to streamline operations and enhance efficiency by automating routine tasks. Preventive asset maintenance focused on proactively identifying and resolving potential IT infrastructure issues, reducing downtime. Predictive analytics leveraged data to provide insights for informed decision-making. The AIDSS prototype's development marked a crucial step towards improving operational efficiency and enabling data-driven decision-making within the ICT department. Rigorous research evaluation methods, including Goal Question Metric and stakeholder feedback using a questionnaire were employed to assess the artefact's effectiveness, usability and impact. Through iteration, continuous improvements and refinements were made to the artefact, considering the unique context and needs of South African higher education institutions. The study contributes to the field by providing a novel and practical solution that enhances decision-making processes, empowers ICT personnel and advances the understanding of AI-enabled decision support systems in the higher education context. In addition, this study engaged in seminal and recent literature and debates on the subject of AI. Thus, the main theoretical contributions were in the generation of knowledge and theory towards the Information Systems discipline. The study established that the artefact would enhance the skills and expertise of ICT personnel; it also provides information that helps users to make decisions effectively. It was also revealed that the system provided appropriate error messages and clear instructions of how to address the errors. In addition, the system successfully predicted and prevented impending ICT issues before they escalated. With these findings, the researcher acknowledges that the artefact contributed to the broader field of AI in higher education, offering practical insights that could guide future research and inform policy making in the context of information systems within academic institutions. The principal objective of this study was to develop an AI-enabled decision-support system; therefore, an artefact was produced at the end of the study.