Machine translation method using part-of-speech information

Abstract

The invention discloses a machine translation method using part-of-speech information, which is characterized in that a training set of bilingual corpora containing corresponding part-of-speech tags is used, and after a translation model is trained, the part-of-speech information is used for assisting machine translation; the translation model comprises a shared encoder, a decoder and a part-of-speech classifier, part-of-speech information is imported in different modes, and the ability of model language understanding and text generation is enhanced; the shared encoder is composed of a source sentence encoder and an auxiliary sentence encoder; the decoder comprises an original multi-head attention module and an auxiliary multi-head attention module. Compared with the prior art, the method has the advantages that the performance of a machine translation model is improved through introduced part-of-speech information, the model is translated on the basis of Transform, the part-of-speech information is imported in two components of an encoder and a decoder in different modes, the translation model can utilize the part-of-speech information to enhance the ability of model language understanding and text generation, and the model language understanding and text generation efficiency is improved. And the machine translation quality is improved.

Description

technical field [0001] The invention relates to the technical field of machine translation, in particular to a machine translation method using part-of-speech information to improve translation quality. Background technique [0002] At present, machine translation technology generally adopts typical machine translation, multi-source machine translation, Transformer model, part-of-speech tagging or BPE algorithm. The goal of a typical machine translation is to translate a sentence in the source language into a sentence in the target language, mark the source language with S, mark the target language with T, and use X for the i-th sentence in a language L (ie S or T) L (i) Indicates that the data set on this language consists of these sentences, expressed as N L is the number of these statements. Each sentence consists of a series of ordered tokens, namely where each symbol comes from the vocabulary V of the language L L , that is, any symbol x i ∈V L . The goal of ...

AI Technical Summary

Problems solved by technology

[0006] The existing technology of machine translation does not fully utilize part-of-speech information, and part-of-speech information is only used for the semantic understanding of the translation model, not for improving the text generation ability of the translation model

Images