An AI-Driven Approach for Advancing English Learning in Educational Information Systems Using Machine Learning

—In current era of globalization, English language learning is important as it has become a global language and helps people to communicate from various regions and languages. For vocational students whose main aim is to get skills and get employed, learning English for communication is important. We here present a proposed framework for learning English language which can become a foundation for a complete Artificial Intelligence (AI) based system for help and guidance to the educators. This study explores the use of diverse Natural Language Processing (NLP) techniques to predict various grammatical aspects of English language content especially focused on tense prediction which lay the foundation of English content. Textual features of Bag of words (BoW) which considers each word as a separate token and Term Frequency –Inverse Document Frequency (TF-IDF) are explored. For both diverse features, the shallow machine learning models of Support Vector Machine (SVM) and Multinomial Naïve Bayes are applied. Moreover, the ensemble models based on Bagging and Calibrated are applied. The results reveal that BoW model input for SVM and Bagging technique using TF-IDF shows optimal results with high accuracy of 90% and 89% respectively. This empirical analysis confirms that such models can be integrated with web or android based systems which can be helpful for learners of English language.