An Intelligent Teaching Aid System for English Linguistics Courses Based on Computer Generated Dialogue System

Abstract

Information technology is of great significance in the curriculum setting and implementation of college English. Traditional college English pays more attention to the extension of the application of multimedia facilities, which is characterized by the problems of untimely updating of content, heavy workload of teachers' preparation and insufficient reinforcement of grammatical context. In this paper, a reinforcement learning-based dialog generation model is proposed based on reinforcement learning theory, which fully considers the one-to-many relationship in dialog data, i.e., there may be multiple reasonable replies for each user input. Then on the basis of the model, an intelligent teaching dialogue system is further developed to realize assisted teaching for English language courses. Through the comparison experiments and case studies of different baseline models in the dataset, it is found that the dialogue generation model proposed in this paper performs better than the other baseline models in various indexes, while the intelligent system can still support about 55% of the users to complete the request when the users reach 700 in the functionality test. Finally, the statistical analysis of the performance in the study case shows that the intelligent dialog generation system proposed in this paper can affect the actual effectiveness of teaching and learning.