Mapping Trends in Artificial Intelligence and Educational Assessment

Abstract

The digital era transformed education, offering promising development opportunities. However, despite having a significant amount of data, education still lacks specific mechanisms for using it to improve learning, teaching and decision-making.  There is a growing body of research advocating for application of AI in education. Advantages of AI come with its ethical concerns related to bias, transparency and privacy. At the same time AI-based assessment is still underexplored in literature. In addition, emerging research trends, links between AI and assessment and existing research communities remain largely unexamined. The paper aimed at exploring the evolving research patterns, link between AI and educational assessment and existing research communities.

This study adopts a bibliometric methodology to analyze the research literature on AI and assessment. Thus, metadata was collected from the Web of Science by Clarivate and Scopus databases over a span of almost 15 years. The obtained data was cleaned, standardized, and combined, resulting in a corpus of 1,465 publications. VOSviewer was used to visualize thematic clusters, author networks, and key areas reflecting current trends in AI in educational assessment. The bibliometric analysis reveals the growing use of machine learning, learning analytics, and intelligent mentoring systems to personalize the educational process and support academic success in modern research.

Keywords: Artificial intelligence, education, assessment,  bibliometric analysis, machine learning.