Exploring the Potential of Artificial Intelligence in Special Education: A Bibliometric Analysis to Support Educators and Students with Autism Spectrum Disorder (ASD)

This full paper describes a complete article in the research category. This work presents a Bibliometric Review that aims to characterize the current scenario of academic research in Artificial Intelligence in education to support educators in the teaching and learning processes in Special Education in different contexts, becoming a tool with potential use in cases of students with Autism Spectrum Disorder (ASD). In this sense, this work aims to contribute to a better understanding of research associated with the area of Artificial Intelligence aimed at educational processes involving students with autism. To do this, we used the EndNote reference manager, an online tool designed to support researchers in conducting Literature Reviews, following the bibliometric protocol proposed by Guedes and Borschiver (2005). From the definition of search strings, (“artificial intelligence”) AND (“autism” OR “ASD” OR “autism spectrum disorder”) AND (“education” OR “teaching”), scientific databases were explored like Web of Science (WoS), Scopus, ERIC, Emerald, Scielo, Portal CAPES, and IEEE, to locate existing studies on the topic, between 2019 and 2024. As a result, 298 articles were mapped and 20 scientific works were selected that address the aforementioned specific theme, written by 89 authors belonging to 50 institutions in 19 countries on four continents, enabling the creation of a significant theoretical basis. By analyzing the information contained in scientific publications in this specific field, it was possible to infer that research is divided into distinct areas of concentration: 1) the exploration of algorithms and data analysis based on Artificial Intelligence for analytical-predictive issues in the field of special education; 2) the use of robots in the teaching and learning processes of students with autism; 3) the development of personalized educational intervention models for learning pedagogical, social and communication skills. The data show that research on the use of Artificial Intelligence in the education of autistic people can still be explored, presenting itself as a promising area for future investigations. The present study facilitates understanding of the current state of knowledge in this area, by identifying gaps in areas that require investigation, thus promoting the development of further research. Finally, the data observed with this research will serve as a basis for the development of future AI projects in the teaching and learning process for autistic students.