Research on Oral Language Evaluation Method of Business English Based on Speech Recognition

In order to meet the demand of people for second language learning, this paper proposes a business English oral evaluation and correction method. Specifically, based on the Transform language model, an end-to-end oral evaluation model based on text priori was adopted to evaluate learners' oral English. Then, in view of the non-standard pronunciation in the evaluation, a two-stage correction model of Normalizing Flow was adopted to correct pronunciation, thereby improving the pronunciation quality of spoken business English. The results show that compared with the traditional mainstream evaluation model, the evaluation performance of the designed evaluation model is significantly better, the response is faster and the evaluation effect is more obvious. The F1 score of the designed evaluation model reaches 0.577 in the evaluation of speech with native accent. In addition, the pronunciation correction model in the evaluation method designed in this paper can preserve the speaker's accent features to the maximum extent on the basis of ensuring the correction effect, and can better help the speaker to learn oral language. It can be concluded that the proposed business English oral evaluation and correction method has good evaluation effect, and can assist users to learn oral language through pronunciation correction, which has certain practical application value.