The field of Data Science has become prevalent across various disciplines, yet its application in the domain of Education remains underdeveloped. Despite a decade of increasing use, a notable absence has been observed in the establishment of a dedicated community, society, and journal for Educational Data Science. The present research endeavour aims to address this dearth of knowledge by means of a state-of-the-art review of Educational Data Science. The paper will (i) provide an overview of the existing definitions of Educational Data Science, and (ii) discuss the specificities of Educational Data Science, particularly concerning data, methods, and ethics. Utilising the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) model, a systematic literature review has been conducted. A comprehensive search strategy has been employed, using keywords to navigate major databases for theoretical and empirical studies, encompassing peer-reviewed articles, books, and institutional reports, without imposing language or time restrictions. A snowballing search has complemented the literature search. The analysis of the records collected has explored the question of whether Educational Data Science is an emerging discipline or an umbrella term, incorporating diverse perspectives from researchers, educators, and practitioners. The findings have substantiated the notion that Educational Data Science is an underdeveloped field of study and that, in the rapidly evolving information age, it represents a burgeoning area of research. A comprehensive understanding of the unique characteristics of Educational Data Science is necessary to ensure effective policy and practice.