Privacy Preserved Reinforcement Learning Model Using Generative AI for Personalized E-Learning
Amutha Prabakar Muniyandi, Balamurugan Balusamy, Rajesh Kumar Dhanaraj, Vijayan Ellappan, S. Murali, Malathy Sathyamoorthy, Lewis Nkenyereye
Abstract
Artificial intelligence algorithms are taking important roleplays in online recommendation models for achieving a high probability of success and these systems are slowly occupying modern learning systems. Modernized learning environments are designed based on personalized E-Learning system, due to the availability of enriched content and flexibility in the learning system. This paper proposed a personalized enriched course recommendation method for an e-learning environment using reinforcement techniques. The proposed method uses an Improved Artificial Bee Colony Optimisation (IABCO) algorithm-based generative AI model for preparing the course recommendations and this recommendation part will act as an Agent in the proposed personalized learning method. The proposed method uses IABCO algorithm for generating enriched course list based on personalized recommendation gather from customers and reinforcement learning model is used to evaluate the suggested course list. The proposed method is experimented with a dataset of online course offering website, which contains 3523 course details and 200 students are taken from various levels of learning maturity. The performance evaluation for the proposed system is measured based on success and accuracy rate of selection from the recommended course list. The average success rate and accuracy for the proposed method is 86.5% and 95.6% compared to the existing AI-based recommendation methods.