Neural machine translation method based on Multi-BiRNN encoding

Abstract

The invention relates to a neural machine translation method based on Multi-BiRNN encoding. Multi-BiRNN encoding is adopted at an encoder end, that is to say that on the basis of using source languagesentences as input sequences, one or more groups of BiRNN are added to encode other input sequences associated with the input sequences; on the basis of Multi-BiRNN-encoded neural machine translation, in the encoding process of a source end, the source language sentence sequences and a dependency syntax tree thereof are considered at the same time, and serialization results of the syntax tree areobtained by means of two different traversal modes respectively and serve as Multi-BiRNN encoding input with the source language sentence sequences; at the output end of each BiRNN, a word is formedin a vector splicing mode. According to the method, vectors obtained by encoding contain more abundant semantic information, the source language sentence sequences and other sequences associated withthe source language sentence sequences are considered at the same time, and the disambiguation function is achieved in the semantic representation process of the source language sentences.

Description

technical field

[0001] The invention relates to a natural language translation technology, in particular to a neural machine translation method based on Multi-BiRNN coding. Background technique

[0002] As a brand-new machine translation method, end-to-end neural machine translation has developed rapidly in recent years. However, end-to-end neural machine translation only uses a nonlinear neural network to convert between natural languages, making it difficult to exploit linguistic knowledge explicitly. How to improve the current framework of neural machine translation, so as to encode and apply linguistic knowledge such as syntactic information to end-to-end neural networks, is a direction worth exploring.

[0003] Usually, end-to-end neural machine translation is based on an "encoder-decoder" framework to learn the transformation rules from the source language to the target language, and the semantic equivalence is described by the vector connecting the encoder and decode...

Images