Research on Digital Business English Teaching Mode under the Background of Big Data Technology

Abstract

Abstract This paper proposes an automatic grammar error correction model based on an encoder-decoder structure, which is combined with a new business English teaching model that can improve the teaching quality of teachers and enable students to get feedback quickly. The model employs a dual encoder structure and extracts source sentence information through an encoder that is based on the attention mechanism. Meanwhile, a fluency-based data augmentation method is also proposed in order to be able to select sentences with different fluency levels based on the fluency ratio between the target sentence and the source sentence. Finally, a Passive-Aggressive algorithm for candidate word selection has been designed, which can avoid generating repetitive and short-biased English sentences. It has been verified that the total grade of the class that incorporates this paper’s model in business English teaching has increased by 9.3 points on average, with the most obvious improvement in business writing by 3 points, followed by translation by 2.8 points. The proposed teaching model in this paper can enhance students’ comprehensive use of business English.