Optimization and implementation of an English MOOC answering system for smart education using LSTM and BERT models

This study aims to design and implement an English MOOC answering system based on intelligent algorithms to improve the efficiency and quality of question-answering services in online education platforms. By adopting machine learning, deep learning, and natural language processing technologies, a system that can automatically generate and match correct answers is built. The results show that the Long Short-Term Memory (LSTM) and Bidirectional Encoder Representations from Transformers (BERT) models perform well in terms of accuracy and recall. Among them, LSTM has a strong ability to solve complex semantic problems, with an accuracy of 92% and a recall of 90%. These experimental results confirm the potential of answering systems based on intelligent algorithms in personalized learning and educational support. The system is designed to support smart education frameworks by enhancing learning adaptability and improving the real-time effectiveness of instructional feedback, in order to better meet the needs of modern smart education. It provides scalable and interactive services aligned with intelligent learning environments, thereby promoting the evolution of data-driven educational platforms.