JAPANESE-VIETNAMESE MACHINE TRANSLATION USES RECURRENT NEURAL NETWORK MODEL

Abstract

Neural network models using deep learning has brought great success when applied in the field of object recognition, speech recognition. For the past three years, this model has been applied to the various tasks of natural language processing such as machine translation, phrase detection, and extraction of words represented by vectors. In this area, machine translation using neural network has initially achieved better results than the traditional linear statistical model. In this paper, we apply neural network model with deep learning technique for machine translation on the task of Japanese-Vietnamese translation and evaluate the results obtained through BLEU metric as 7,7 point. We also point out the shortcomings to be solved to improve the quality of translation system in the future.