Method and device for translation

Abstract

The embodiment of the invention discloses a method and device for translation. The method comprises the steps of obtaining a source language text input by a user and an initial syntax tree input by the user; inputting the source language text and the initial syntax tree into a syntax analysis network of the translation model to obtain an optimized syntax tree output by the syntax analysis network;and inputting the source language text and the optimized syntax tree into a syntax translation network of a translation model to obtain a target language text which is output by the syntax translation network and conforms to a target syntax structure, the target syntax structure being a syntax structure indicated by the optimized syntax tree output by a syntax analysis network. According to the method, the pertinence of the output target language text can be improved, so that the output target language text better conforms to the style of the translation expected by the user, and the requirements of different translation scenes are better met.

Description

technical field [0001] The present disclosure relates to the field of computer technology, in particular to the field of translation technology, and in particular to a method and device for translation. Background technique [0002] The style of translated sentences generated by traditional translation models depends on the translation style of sentences in the used training corpus. For example, most of the training corpora used for translation are written expressions, which are characterized by more rigorous words and more complex sentence patterns. [0003] In practical applications, for some spoken languages, such as translation of tourism scenes, simultaneous interpretation, etc., the translated text is usually required to be relatively concise, and the translation that does not produce long-distance dependencies as much as possible. Based on this requirement, it is necessary to forcibly generate a translation with a specified grammar, so that the translation can meet t...

AI Technical Summary

Problems solved by technology

[0005] However, the existing methods for generating translations with specified grammars require the collection of corresponding training corpus, and the acquisition of training corpus usually requires a certain amount of labor cost for labeling

In addition, using the translation model after incremental training of the training corpus, there is no guarantee that the translation of the desired grammatical structure will be produced when translating a specific sentence

Images