Intelligent Language Acquisition Model for Online Student Interaction with Educators Using 6G-Cyber Enhanced Wireless Network
Lijuan Zheng
DOI: https://doi.org/10.1007/s11277-024-11197-x
IF: 2.017
2024-06-18
Wireless Personal Communications
Abstract:Deep learning have created personalized, adaptive, and secure e-learning environments in the ever-changing educational landscape. Privacy, bias, and data may limit e-learning systems. Students can experience more agile, personalized, and secure e-learning with Cloud-to-Thing Continuum. In this research, we proposed Cloud-to-Thing Continuum, an innovative Early Warning System that automates individualized recommendations and tracing in a safe AI-assisted adaptive Smart Classroom and predicts and improves student performance. The initial focus is FL-based Authentication. A secure and private authentication system for Course Instructors (CIs) and students is needed. This unique solution registers user credentials with a central server and produces a global model using Cycle General Adversarial Network. In the adaptive e-learning environment, Intelligent Software Agents (ISAs) play a crucial role in assessing weekly student engagement. They capture students' interactiveness, academic emotions, and behaviors, and monitor CIs teaching styles. ISAs use the Threshold-based Shannon Entropy method to classify and evaluate these factors, categorizing students into engaged or not engaged clusters through Majority Voting-based Multi-Objective Clustering. This approach improves adaptive e-learning by assessing student participation based on academic and behavioral profiles. Leveraging weekly engagement status, demographics, and an innovative DRL-based early warning system, using Improved Dueling Deep Q Network, we predict not-engaged students' performance, categorizing them into three classes: dropout risk, lower final exam scores, facilitating personalized recommendations. The multi-disciplinary ontology graph, derived from student performance, drives automated personalized recommendations using Att-CapsNet. It queries relevant data to offer tailored solutions, such as scholarships or interactive teaching methods, while ongoing performance tracking enhances student engagement. All data is securely stored in a blockchain using Light Weight Encryption Algorithm encryption.
telecommunications