Dan Zhang and Omer Hassan Ali Mahfoodh
HMI, AI education systems, chaos-infused nutcracker-mutated temporal gated recurrent unit with multilayer perceptron(CHN–TGRU–MLP), personalised learning.
The proposed framework focuses on enhancing human–machine interaction (HMI) in AI-driven education systems through real- time feedback mechanisms. By utilising sophisticated data-driven AI modelling, the system adapts educational content based on learner performance, delivering personalised feedback. The core architecture, chaos-infused nutcracker-mutated temporal gated recurrent unit with multilayer perceptron, combines chaos-based optimisation and adaptive learning to enhance HMI. The CHN-TGRU handles variability and captures sequential learning patterns, while the MLP dynamically adjusts feedback by predicting learner responses. The system collects multimodal learner data, including text, voice commands, and response times. Text data is processed by removing stopwords, and voice data undergoes normalisation with mel-frequency cepstral coefficients (MFCC) extracted to capture spectral and phonetic features. A reinforcement learning (RL) agent generates real-time feedback based on predicted learner performance. Simulations of the system reveal that the CHN–TGRU–MLP method outperforms baseline models, achieving 0.91 accuracy, faster response times, higher engagement, and improved task completion rates. The integration of CHN–TGRU–MLP and RL- based feedback demonstrates significant advancements in adaptive learning, improving learner performance, engagement, and retention. This scalable framework offers an efficient and personalised approach to intelligent human–machine interactive learning.
Important Links:
Go Back
