A text translation method and related device

Abstract

The invention discloses a text translation method, which comprises the following steps: obtaining a text sequence to be translated; Performing encoding processing on the text sequence to be translatedto obtain a first implied state sequence; Acquiring a first state vector; Generating a second hidden state sequence according to the first state vector and the first hidden state sequence; Generatinga context vector corresponding to the current word according to the second implied state sequence and the first state vector; According to the context vector, the first state vector and the first target word, determining the second target word, wherein, the first target word is the translation result corresponding to the previous word, and the second target word is the translation result corresponding to the current word. The invention also discloses a text translation device. The embodiment of the invention introduces a context vector obtained by decoding in the process of encoding the textsequence to be translated corresponding to the source language text, thereby enhancing the representation of the text sequence to be translated and improving the translation quality.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to a text translation method and a related device. Background technique [0002] In recent years, the encoder-decoder framework has achieved outstanding results in text processing tasks, including machine dialogue, machine question answering, and machine translation. In the machine translation project, different languages ​​can be translated. For example, if the input sequence is an English sentence, then the output can be the Chinese translation result of the English sentence. [0003] At present, the process of translating using the encoder-decoder framework as a translation model is as follows: first, the source language sentence is converted into a vector representation, and then the vector representation sequence is input to the encoder, and the intermediate vector is obtained after encoding. Finally, the The decoder decodes the intermediate vector to generate a translat...

AI Technical Summary

Problems solved by technology

[0004] However, using the encoder-decoder framework for translation can achieve the purpose of translation, but the translation quality is not high, especially for long sentences, it is more prone to translation deviations, resulting in poor translation results

Images