Abstract
Neural machine translation is a new approach to machine translation that has shown the effective results for high-resource languages. Recently, the attention-based neural machine translation with the large scale parallel corpus plays an important role to achieve high performance for translation results. In this research, a parallel corpus for Myanmar-English language pair is prepared and attention-based neural machine translation models are introduced based on word to word level, character to word level, and syllable to word level. We do the experiments of the proposed model to translate the long sentences and to address morphological problems. To decrease the low resource problem, source side monolingual data are also used. So, this work investigates to improve Myanmar to English neural machine translation system. The experimental results show that syllable to word level neural mahine translation model obtains an improvement over the baseline systems.
Highlights
In recent year, neural machine translation (NMT) has been proposed and got the superior performance results in many language pairs
NMT models based on the encoder and decoder framework, it consists of one encoder RNN (Recurrent Neural Network) and one decoder RNN
Rests of the paper are organized as follows: section 2 describes the related work; section 3 describes about Myanmar Language; section 4 describes the model of attention-based NMT; section 5 describes the experimental setting and section 6 is the conclusion and future work
Summary
Neural machine translation (NMT) has been proposed and got the superior performance results in many language pairs. Neural machine translation (NMT) with the attention-based encoder-decoder framework has achieved significant improvements in translation quality of many language pairs. The small scale of Myanmar-English parallel corpus is not sufficient to build a neural translation model. Attention-based neural machine translation on the Myanmar to English machine translation with this corpus. Based on the attention-based NMT, we further apply character to word level and syllable to word level attention-based neural machine translation on the Myanmar to English machine translation with the same corpus. By using source-side monolingual data for low-resource languages, it is better to boost the performance of the low-resource Myanmar to English neural machine translation model. The BLEU of source side monolingual data gets the 28.51 and obtains an improvement of 6.43 BLEU over the word to word level neural machine translation system. Rests of the paper are organized as follows: section 2 describes the related work; section 3 describes about Myanmar Language; section 4 describes the model of attention-based NMT; section 5 describes the experimental setting and section 6 is the conclusion and future work
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
