AI Adaptive Math Learning Platform: Creating Personalized Educational Experiences for Visually and Hearing Impaired Students

This project is an educational platform integrating machine learning, generative AI, and explainable AI technologies. It addresses the math learning needs of visually and hearing impaired students by providing personalized Pythagorean theorem teaching and intelligent tutoring, solving the shortcomings of traditional education for students with special needs, and promoting inclusive education.

The traditional education system has obvious shortcomings in catering to students with special needs: visually impaired students struggle to access visual math content, while hearing impaired students are confused by audio explanations; most platforms use a "one-size-fits-all" content delivery method, ignoring differences in learning speed and cognition, leading to significantly lower learning outcomes for students with special needs in STEM fields compared to ordinary students.

Core Technology Architecture

1. Learner Profile and Classification System: Collects information through pre-assessment, uses machine learning to classify students into slow/advanced learners, and combines disability type to lay the foundation for personalization
2. Multimodal Content Generation Engine:
   • Visually impaired support: Audio explanations, tactile/auditory-friendly materials, voice chatbots, adaptive prompts
   • Hearing impaired support: Animated visual explanations, interactive modules, visual anchor paths, visual quizzes
3. Integration of Generative AI and Explainable AI: Generative AI generates personalized content in real-time, while XAI ensures interpretability; LSTM tracks learning behavior patterns to predict difficulties

Closed-Loop Learning Process

1. Pre-assessment → 2. Intelligent classification →3. Content personalization →4. Adaptive learning →5. Post-assessment →6. Feedback recommendation →7. Continuous tracking

This project promotes inclusive education: through AI personalization, it enhances the participation and depth of understanding in math learning for students with special needs, boosts their learning confidence and independent learning ability; it demonstrates the potential of AI in addressing educational inequality, providing a feasible technical path for building a fair learning environment.

The AI adaptive math learning platform is a modern interpretation of the ancient concept of "teaching according to individual needs". It combines advanced technology with humanistic care, indicating the direction of future educational technology: not to replace teachers, but to empower every learner, regardless of the challenges they face.

The project team plans to expand functions:

• Cover a wider range of math concepts
• Integrate sign language video explanations
• Real-time speech-to-text and text-to-speech
• Multilingual learning modules
• Advanced AI cognitive tutoring capabilities