Abstract:

Digital transformation and technological advancements are rapidly propelling significant progress in AI, which is becoming increasingly transformative in education. This research aims to examine studies on AI across K-12, higher, and graduate educational levels and in both remote and in-person settings.

We employed a bibliometric analysis in this study. Using the PRISMA methodology, we identified 465 publications on the Web of Science and included 220 articles from 105 journals in our research scope, spanning from 2003 to 2023. The annual publication growth rate is 22.6%. The average age of the documents within the dataset is 3.02 years, with an average of 10.55 citations per document, totaling 7,947 references. Our dataset includes contributions from 657 authors, 49 of whom authored single-authored documents. Collaboration among authors is evident, with an average of 3.23 co-authors per document. International co-authorships account for 19.55% of collaborations. In 2022, we had the most publications, with 69 records comprising 31.364% of the dataset.

The top three contributing journals are Mobile Information Systems, Frontiers in Psychology, and Journal of Intelligent & Fuzzy Systems.

China leads in scientific production with 88 publications, followed by Saudi Arabia with 20 publications and the United States with 14 publications.

Publications was clustered under five categories based on the author keyword co-occurrence analysis.

The first one is Artificial Intelligence, Machine Learning, and Data Analytics cluster. It covers adaptive learning, personalized education, sentiment analysis, predictive modeling, and intelligent tutoring systems to enhance the learning experience, tailor instruction to individual needs, predict student performance, engagement, and learning outcomes, and automate various educational processes.

The Online, Distance, and Digital Learning cluster refers to educational approaches that utilize AI technologies for remote instruction, including e-learning, SPOCs, MOOCs, virtual classrooms, and mobile learning.

The Educational Technology cluster includes AI-based digital teaching, learning, and educational administration applications. These tools enhance engagement, facilitate communication, streamline administrative tasks, and provide personalized learning experiences, including gamification and smart classrooms.

The Pedagogical Approaches and Teaching Methods cluster encompasses diverse AI-based instructional strategies that facilitate learning, including flipped classrooms and blended learning, and increase student engagement, performance, perceptions, satisfaction, and self-regulation. These methods provide appropriate support and foster positive learning experiences.

The Challenges and Issues category addresses concerns about equity, accessibility, well-being, and learners' privacy in AI-based settings. Based on the results, while most reviewed studies concentrate on higher and adult education, AI literacy is significant at K-12, primary, and secondary school levels.

In conclusion, the research underscores AI's growing importance in education and provides insights into trends and collaborations in the field. As AI reshapes teaching and learning, addressing challenges and ensuring access to these technologies is crucial. Future work should promote AI literacy at all educational levels for inclusive, transformative learning.

Keywords:

Artificial Intelligence, Bibliometric Analysis, Science Mapping, Education, Learning.