AN INTELLIGENT FRAMEWORK FOR REAL-TIME FEEDBACK IN MOBILE LEARNING ENVIRONMENTS FOR CHILDREN

Abstract

This paper describes an intelligent framework for providing real-time feedback in mobile learning environments for kids. The framework employs Edge-Fog-Cloud computing methods to designed a safe, low latency, decision-making, and compliant with child data protection standards. The framework’s center piece integrates two components, knowledge state estimation using Long Short-Term Memory (LSTM) networks and emotional state using physiological and behavioral. Every generation of feedback incorporated safe decision-making policies to guarantee ethical and age-appropriate feedback. Also, a multi-objective optimization framework balances the trade-off between accuracy, latency, engagement, and the efficient use of resources in the system to optimal resource utilization. The system performs better than baseline methodologies, including CNNs and SVMs, in engagement detection, as evidenced by reduced mean squared error and latency, as well as improved F1 scores. The system for mobile learning environments for children and the associated methodologies seamlessly scale, are adjustable, and preserve user privacy.