A structured text translation method and device

Abstract

Embodiments of the present invention provide a method and device for translating structured text, including: removing the structural markup of the target structured text to be translated to obtain the target text; inputting the target text into the trained text translation neural network model In this method, search and translate the translation candidates of the target text according to the phrase search space to obtain the target translation text and alignment information; according to the alignment information, carry out structured mark recovery processing on the target translation text to obtain the target structured Translate text. In the embodiment of the present invention, by removing the structured markup of the structured text, the text without the structured markup is translated through the neural network model based on the phrase search space, and the translated text is restored to the structured markup to obtain a structured translation Text, which realizes the translation of structured text through the neural network model.

Description

technical field [0001] The invention relates to the technical field of machine translation, in particular to a structured text translation method and device. Background technique [0002] In recent years, the rapid development of neural network machine translation technology has significantly improved the quality of machine translation. Furthermore, the improvement of the quality of machine translation has also made it widely used in real life. [0003] Although neural machine translation is excellent in translating plain text, it still cannot be well applied to translating structured text, because the translation of structured text needs to satisfy structural constraints, for example, the source end is between a pair of HTML tags The translation corresponding to the content of the target must also be included between the same pair of HTML tags, but because in the existing neural machine translation, there is no structured text corpus for training the model for structured t...

AI Technical Summary

Problems solved by technology

[0003] Although neural machine translation is excellent in translating plain text, it still cannot be applied to translating structured text well, because the translation of structured text needs to satisfy structural constraints, for example, the source end is between a pair of HTML tags The translation corresponding to the content of the target must also be included between the same pair of HTML tags, but because in the existing neural machine translation, there is no structured text corpus for training the model for structured text translation; and , neural machine translation lacks explicit alignment information and cannot add structural constraints, making it difficult for existing neural machine translation to translate structured text

Images