Explainable Models for Predicting Academic Pathways for High School Students in Saudi Arabia

The science of data mining has contributed considerably to the education sector. However, most educational data mining (EDM) studies have focused on predicting the future performance of students and detecting at-risk students to provide early targeted interventions. A few studies used machine learning techniques to predict the future academic pathways for degree students. However, there are limited studies that use the datasets of high school students to predict future pathways. Moreover, such studies are yet to be conducted using datasets produced in Saudi high schools. Therefore, researchers that work in the education sector have focused on EDM and are eager to apply advanced computer science methods to upgrade old administrative systems. Furthermore, the education sector is a rich field with data that can be exploited to improve and enhance educational management systems and accelerate digital transformation. In this study, we explore the applications of EDM and review the algorithms used by other researchers in this field. We applied supervised machine learning classifiers to educational datasets collected from high schools in Saudi to predict the future academic pathways of students and identify the essential factors that affect them. This study contributes to the literature by developing a predictive model for students in Saudi high schools and detecting critical features that affect future academic careers of the students. Furthermore, we used explainable artificial intelligence to interpret the best model and enhance its transparency.