An integrated Business Intelligence Model to Enhance Learning Processes in Higher Education Institutions

Abstract

Business Intelligence (BI) and learning analytic (LA) are drawing substantial attentions recently, and being one of the most competitive advantageous-themes areas in universities nowadays. The revolution of ICT and the vast amount of data that has been generated accordingly, makes it significant to gain the maximum benefits of these datasets, in proportion to retention, students' success as a key research topic and students’ satisfaction. And to improve individual performance, that leed to improve the overall organization's performance, increase profitability, novel insights and stronger innovations. Although BI systems have a great impact on strategic, informed and high-quality decisions based, there is a lack of evident based practical guidance, on how effectively deploying LA to improve learning outcomes and students’ success, as a big challenge in HEIs. The objective of this research, is to fill the void in literature by developing and validating a framework. The model is expected to integrates BI software solutions, LA and students’ performance. Predictive analytics is used for analyzing useful knowledge from the student learning experience, to improve learning outcomes for the students and the society. A student satisfaction survey via a questionnaire is conducted, to capture student requirements, and for well understanding of their educational needs, as well as individual learning characteristics. The understanding of quality office requirements, gathered by interviewing quality affairs dean, studying university’s documents and distributing questionnaire to academic and administrative s-taff member’s representatives, for many key regions. Researcher conducted a systematic literature review (LR), in LA field action researches, design and development LA models, universities’ BI initiatives and big data generated in HE as a big challenge. Using design and development research methods, a novel but practical model was constructed. The model is validated, deploying a real students’ dataset of “1034” undergraduate active students (students’ mean age 19.5, 65% females), studying for degree of computer science in Computer Science and Information Technology College in SUST, one of the public universities in Sudan. Starting with extracting the variables, from heterogeneous resources (student information system, students’ grading, enrollment and biography information), to populate warehouse systems, which used for performance measurement and decision support by analyzing facts produced, using Tableau BI software tool. The findings explored that, how technology captured data of students’ performance for prediction, to identified at-risk students, for the purpose of consulting and withholding them before being drop out. Findings also investigate a profound domain knowledge about students and their context, and effectively differentiate the overall performance per college/departments per years/semester, giving implication of the retention, completion rate for graduated students and to detect defects to be corrected using and implementing effective informed decisions. Finally, a comprehensive evaluation is conducted, by surveys and interviews, determining the efficiency and influence of the proposed framework. This research identifies five main phases of this integrated framework: Data warehouse population, Data analytics, Visualization, Operational insights and assessment data phase. Each stage involves several key fundamental factors.