Educational Data Mining in Prediction of Students’ Learning Performance: A Scoping Review

Students’ academic achievement is always a target of concern for educational institutions. Nowadays, the rapid development of digital transformation has resulted in huge amounts of data being generated by Learning Management Systems. The deployment of Educational Data Mining (EDM) is becoming increasingly significant in discovering ways to improve student learning outcomes. Those approaches effectively facilitate dealing with students’ massive amounts of data. The purpose of this review is to evaluate and discuss the state-of-art EDM for predicting students’ learning performance among higher education institutions. A scoping review was conducted on twelve peer-reviewed publications that were indexed in ACM, IEEE Xplore, Science Direct and Scopus between 2012 and 2021. This study comprehensively reviewed the final inclusion literature on EDM in terms of tools, techniques, machine learning algorithms and application schemes. We have found that WEKA (tool) and classification (technique) were chosen in most of the selected studies carried out in their EDM settings. This review suggested that Tree Structured algorithms as supervised learning approaches can better predict students’ learning performance, as it has been validated in several comparative analyses of other algorithms. In the present study, we demonstrate a future trend toward improving the generalizability of prediction models that can deal with a diverse population and the predictive results can be easily interpreted and explained by educators in the general market.