Development of an AI-driven Personalized English Learning Platform Based on Machine Learning

Introduction:English language acquisition is essential for global communication, academic achievement, and professional development. However, many existing digital learning platforms lack true personalization and fail to adapt effectively to individual learner profiles, creating a need for fully AI-driven adaptive English learning systems.Objective:The objective of this research is to develop an AI-driven personalized English learning platform using machine-learning techniques to enhance learner engagement, performance prediction, and adaptive content recommendation.Method:Learner interaction data from 2,000 users, including performance scores and engagement metrics, were collected and preprocessed using Min–Max normalization. Semantic and behavioral features were extracted using TF-IDF and statistical embeddings. A Modified Sea Lion Optimizer–tuned Random Support Vector Machine (MSO-RSVM) model was employed for learner performance classification and personalized content recommendation.Results:Experimental evaluation conducted using Python 3.10 demonstrated that the proposed MSO-RSVM model achieved a prediction accuracy of (98,18%)[H2.1]outperforming baseline models such as CNN, CN