A Chinese-English Machine Translation Model Based on Deep Neural Network

摘要

In this paper, the neuro-machine translation model is studied. Firstly, the encoder-decoder framework used in neuro-machine translation is introduced. Then, the neuro-machine translation models based on RNN, LSTM and GRU are constructed respectively according to different network structures. Aiming at the problem of dealing with long-distance dependence, attention mechanism is integrated into translation model, so the preprocessing module, encoder-decoder framework and attention module of the system are also adopted. Furthermore, a translation model based on bidirectional GRU is proposed to improve the translation performance by enhancing the source language context information. A comparative analysis of the above translation models shows that the proposed translation model is effective.