Bibliometric analysis of research patterns in AI based inclusive pedagogy in schools using classic scientometric laws

Abstract

This study analyses patterns of research and writing activity on the role of artificial intelligence in inclusive pedagogy across school education contexts through bibliometric analysis, focusing on classical laws of scientometrics. A total of 2,490 research publications published between 2016 and 2025 from the Dimensions AI database were identified using a comprehensive search string on the topic. The dataset was analysed to uncover trends in country and university contributions, author collaborations, and themes in abstract co-occurrences. It was then examined for author productivity, journal productivity, frequency patterns of keywords in abstracts, and citation analysis of authors, countries, institutions, and sources. The research output has notably increased in recent years. Globally, China recorded the highest publication count within the dataset, followed by the USA and Australia. in this field. A systematic analysis revealed that the dataset aligned with Lotka's Law to some extent, strongly adhered to Bradford's Law, and exhibited Zipfian behaviour. We identified eight thematic clusters ranging from the ethics of AI adoption, pedagogy in curriculum design, and equity considerations to policy innovation. Our citation analysis revealed two complementary forms of influence, illustrating collaborative centrality linked to both research areas and the impact of individual studies.