Talent Training Model of Business English Translation Based on the Graphical Teaching Method

In order to improve the effect of teaching business English translation and realize the teaching goal of translation talent training, this paper puts forward a pronunciation automatic evaluation method based on phonetic model perception. Combined with expert evaluation method, the method can complete English translation pronunciation error tagging by measuring the similarity between the standard speech and the test speech. Constructing a text corpus of HAW speech automatic evaluation, using Viterbi decoding method designs the phonological similarity judgment method, the text corpus is decomposed into phonetic sequence, and then we can obtain the similarity after the use of training phonetic sequence and test sample data are compared. The teaching experiment is carried out to verify the method, the test results show that the test method finally can output speech tagging errors in the form of graphical results, which provides a visual material for the graphical teaching of English translation.