A case study on the development of AI-based reading assistive materials for on-site needs of Korean reading materials for textbook text adaptation of KSL learners

The purpose of this study is to develop customized textbook-based reading materials for Korean as a Second Language (KSL) students using generative AI and to explore its educational applicability. A survey was conducted with 36 public school teachers regarding the development of customized reading materials, and the revisions and satisfaction levels concerning reading material drafts generated by ChatGPT were analyzed. The findings indicate that teachers recognize the necessity of textbook-based reading materials to support KSL students in their academic studies,emphasizing the importance of text difficulty adjustment and the incorporating images. Teachers revised AI-generated reading texts by applying principles of simplification and elaboration at the lexical and sentence levels and made additional modifications, such as re-editing based on the original textbook content or adding supplementary explanations, to achieve learning objectives. While teachers showed moderate to high satisfaction with AI-generated text materials, they assessed that the image resources required improvement. These results suggest that generative AI can serve as an efficient tool for developing customized materials for KSL students,but the production of high-quality materials necessitates enhanced teacher proficiency in AI utilization and appropriate educational interventions. This study provides foundational research on leveraging advanced technologies for the development of tailored, curriculum-linked educational materials for KSL students, establishing a practical basis for implementing individualized education with AI in future educational settings.