An Attention-Based Text Understanding and Automated Question Answering System for Smart College English Classrooms

The rapid development of Artificial Intelligence (AI) and Natural Language Processing (NLP) technologies has significantly transformed educational environments, enabling more personalised, interactive, and intelligent learning experiences. Traditional classroom methods often face limitations like inadequate individualised attention, delayed feedback, and difficulty in accurately assessing comprehension levels. To overcome these challenges, this research presents the design and implementation of a text understanding and automatic Question Answering (QA) system for college English intelligent classrooms, leveraging an attention mechanism to improve comprehension and response accuracy. A scalable Transient Search (STS)-driven Attention-enriched Recurrent Neural Network (STS-Att-RNN) is used to identify critical information within a text and automatically generate accurate answers to student queries. The English Classroom QA Dataset is obtained from Kaggle. Text cleaning and tokenisation are used to preprocess the data. Features are extracted by employing the Term Frequency and Inverse Document Frequency (TF-IDF) approach. After the text is cleaned and tokenised, it is represented as informative TF-IDF features that are fed to the suggested STS-Att-RNN model for text understanding. The attention mechanism helps capture contextual relationships in the college English texts, whereas the STS optimisation ensures effective tuning of model parameters for providing accurate answers to student questions. The attention mechanism enables the QA system to understand text deeply and generate accurate answers in a college English intelligent classroom setting. Findings show the efficiency of the suggested model by achieving high accuracy (98.2%), precision (95.5%), recall (95.8%) and [Formula: see text]1 scores (97.3%) in answer generation, with reduced response time compared to traditional QA models. The results indicate that the attention-driven approach not only improves the accuracy of automatic answers but also facilitates an adaptive and interactive learning environment, thereby supporting personalised education. This system provides a practical and scalable solution for intelligent English classrooms, bridging the gap between automated QA technology and effective teaching practices.