Artificial intalligence in special education: a systematic review

Abstract

The role of techniques involving Artificial Intelligence (AI) has been becoming increasingly important in educational settings. This study aims to reveal the recent trends in research into artificial intelligence in special education by using the systematic review method. Across the 29 studies published between 2008 and 2020 that are reviewed here, the majority are articles on quantitative research carried out in the United States. In terms of learning content, most of the studies are about skill development, focusing on cognitive and affective factors. Moreover, the research was carried out in school settings with learners from various backgrounds. The purposes for and methods of using Al were examined and it was found that software-based methods are more common. The primary disability type examined in the articles covered in this review were disorders on the autism spectrum. There is a tendency toward technical models rather than educational models. ANN and SVM are the most used technical theoretical infrastructures. This study yielded interesting results both on the evolution of Al in special education over the years and on its future development. This paper ends with suggestions developed in light of these studies for the use of artificial intelligence in special education.