Development of an AI-Based Digital Textbook Prototype for Students with Physical Disabilities

Abstract

This study aimed to present a concrete development model of an AI-based digital textbook for students with physical disabilities, grounded in special education teachers’ needs and design principles derived from prior research. A Design-Based Research (DBR) methodology was employed, and prototypes were developed and evaluated through two iterative cycles. At each stage, the process of design, expert review, and evaluation was conducted to derive the final version. In the first cycle, an initial prototype was created that incorporated pre-diagnostic assessment, level-differentiated learning activities, real-time feedback, and evaluation functions, and was revised based on expert feedback. In the second cycle, improvements were made by refining diagnostic functions and adjusting task difficulty. The final prototype integrated pre-, in-, and post-class learning functions, providing personalized pathways, adaptive content delivery, and real-time analytics dashboards. These findings demonstrate the potential of AI-based digital textbooks to support customized learning tailored to students with physical disabilities. Unlike existing assistive digital materials, this study presents an empirically validated design model that connects AI-driven adaptivity with Universal Design for Learning (UDL) principles, offering practical implications for inclusive education. The originality of this work lies in bridging AI-based personalization with special education contexts, establishing foundational design principles for future development across diverse disability types.