LexiLearn: Adaptive AI-Driven Virtual Reality for Multisensory Literacy Instruction and Pedagogical Support

Abstract

We present LexiLearn, an AI-driven, immersive virtual reality intelligent tutoring system designed for foundational literacy acquisition. The system integrates real-time learner performance analytics, adaptive content sequencing, and multimodal interaction to provide personalized spelling and phonics practice within an embodied virtual environment. An AI engine continuously models learner proficiency by tracking accuracy, error patterns, and response timing, dynamically adjusting task difficulty, word selection, and repetition schedules to maintain an optimal challenge level. Generative AI produces categorized relevant visual imagery to reinforce word-meaning associations, while phoneme–grapheme interactions are synchronized with multisensory haptic feedback delivered via haptic gloves. An AI-enhanced teacher dashboard translates learner model outputs into interpretable pedagogical insights through conversational AI, including detecting misconception patterns and recommending classroom interventions. This closes the loop between automated adaptation and human instructional decision-making, positioning the system as a hybrid intelligent tutoring and teacher-support platform.