A Real‐Time English Knowledge Recommendation Framework Integrating ITL Networks and TinyML

ABSTRACT With the rapid growth of mobile learning and edge intelligent devices, real‐time and personalized English knowledge recommendation has become a key requirement in intelligent education. However, traditional recommendation models often suffer from high computational complexity, large model size, and high latency, making them unsuitable for resource‐constrained edge environments. To address this challenge, this paper proposes a real‐time English knowledge recommendation framework that integrates Information‐Theoretic Learning (ITL) networks with Tiny Machine Learning (TinyML). The ITL‐based feature extraction module captures nonlinear correlations between user learning behaviors and knowledge points by maximizing mutual information, enhancing feature representation capability. Meanwhile, a TinyML‐based multi‐scale model compression strategy significantly reduces model size and computational complexity while maintaining high recommendation accuracy. Experimental results demonstrate that the proposed framework achieves superior performance in recommendation accuracy, response latency, and energy efficiency, providing an effective solution for deploying intelligent English knowledge recommendation systems on edge devices.