Development of an Integrated AI-Based Learning Disability Identification and Support Application

Abstract

This study presents the development of an Integrated AI-Based Learning Disability Identification and Support Application designed to assist teachers in the early detection and support of students with learning difficulties, particularly dyslexia, dysgraphia, and dyscalculia. Utilizing machine learning algorithms and AI APIs, the application analyzes student performance data including reading speed, writing samples, quiz performance, and task completion time to identify patterns indicative of learning disabilities. The system features an offline-capable AI analysis module, a performance dashboard, a chatbot-based assessment assistant (KUGI), and a local database. Developed using the Waterfall SDLC model, the application integrates React Native (Expo Go) for cross-platform mobile development and Node.js for backend processing. The study employed developmental research design to systematically create and evaluate the proposed system. Planned evaluation metrics include accuracy, precision, and recall of the AI detection module. Results indicate that the application can provide timely, data-driven insights to support inclusive education in resource-constrained environments, particularly within the Philippine educational context.