Enhancing Cognitive Skills in Students with Moderate Intellectual Disabilities: Effectiveness of an Immersive Adaptive AI Educational System Integrating Natural Language Processing and Multi-Sensory Interaction

Students with moderate intellectual disabilities face considerable challenges when engaging with traditional educational and training approaches, making the learning process demanding for both learners and their instructors. Against the backdrop of rapid advancements in artificial intelligence and natural language processing (NLP), cognitive computing has proven highly effective in creating interactive learning environments that respond dynamically to the individual needs of students with disabilities. This study sought first to identify the specific difficulties experienced by students with moderate intellectual disabilities aged 12 to 17 years, and then to develop an educational software program grounded in cognitive computing and natural language processing technologies. The program incorporates automatic speech recognition (ASR), tactile interaction, and adaptive response generation delivered through an interactive educational character, with the aim of teaching core skills such as measuring sizes and lengths, recognizing colors, and performing basic arithmetic operations. Employing a quasi-experimental design, the study involved two groups: an experimental group and a control group, each consisting of seven students whose IQ scores ranged from 55 to 70. The findings demonstrated the software’s clear effectiveness in enhancing cognitive performance, with correct response rates in the post-test reaching 85% - 90% in favor of the experimental group.