Machine translation method and device based on translation memory bank

Abstract

The invention provides a machine translation method based on a translation memory bank. The method comprises the following steps: searching a corpus original text with the highest similarity with a to-be-translated original text and a translation of the corpus original text from the translation memory bank; comparing the to-be-translated original text with the corpus original text to obtain a difference part different from the to-be-translated original text in the corpus original text; mapping the difference part to a translated text of the corpus original text, and replacing the translated text mapped by the difference part in the translated text of the corpus original text with a mask; taking the replaced translated text of the corpus original text and the to-be-translated original text as the input of a machine translation model, and outputting the translated text of the to-be-translated original text; wherein the machine translation model is obtained by taking a translation original text sample as a sample and taking a translation corresponding to the translation original text sample as a label for training. According to the method, the translation of the to-be-translated original text and the translation of the corpus original text are combined for translation, so that the translation efficiency can be improved, the translation cost can be reduced, and the translation accuracy can be improved.

Description

technical field [0001] The present invention relates to the technical field of machine translation, in particular to a machine translation method and device based on translation memory. Background technique [0002] The translation memory database is the bilingual corpus generated and retained by the translator during the translation process, and it is usually data with relatively high translation quality after manual proofreading. Due to the limited corpus in the translation memory, it is very likely that the exact same corpus as the current text to be translated cannot be retrieved from the translation memory, so that the translation of the current text to be translated cannot be obtained directly from the translation memory. [0003] Translation memory can be used to assist the current translation task. The existing method is to retrieve the corpus similar to the current text to be translated from the translation memory, and present the corresponding translation to the t...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a machine translation method and device based on translation memory, which is used to solve the time-consuming and labor-intensive defects in the prior art when translators check and edit translations of similar corpus, and realize automatic translation of texts to be translated based on translation memory translate

Images