Software-Defined Learning: Innovation in College English Education

Students’ practical English application skills and communication abilities are the two main goals of college English education. However, a lot of students continue to prioritize exam success over the development of thorough language proficiency, especially in spoken English. To address this gap, this paper introduces a novel educational model, Software Defined Learning (SDL), which leverages Artificial Intelligence (AI) and Machine Learning (ML) techniques to adapt learning resources and optimize English learning experiences. A novel Weighted Spider Monkey Optimizer Refined Decision Tree (WSMO-RDT) is applied to evaluate students’ overall English proficiency. In college English education, data are collected through student performance records, including exam scores, oral assessments, and class participation. Preprocessing involves normalizing scores to ensure comparability across different assessments and handling missing or incomplete data. By optimizing the RDT model’s hyperparameters, WSMO and RDT can improve the model’s accuracy in predicting students’ English competence. Based on each student’s performance, this hybrid method aims to offer more individualized and flexible learning pathways. The proposed WSMO-RDT method achieved better outcomes when compared with other traditional techniques. The proposed model, combined with SDL and AI, has revolutionized college English education by providing an F1-score of 96.8%, recall of 95.9%, accuracy of 95.6%, and precision of 95.8% for personalized and adaptive learning experiences. This approach enhances English proficiency beyond exam-focused methods, demonstrating the potential of AI and ML. Future research could extend the model to other educational fields, incorporate real-time feedback, and improve student engagement.