Abstract
Existing Sequence-to-Sequence (Seq2Seq) Neural Machine Translation (NMT) shows strong capability with High-Resource Languages (HRLs). However, this approach poses serious challenges when processing Low-Resource Languages (LRLs), because the model expression is limited by the training scale of parallel sentence pairs. This study utilizes adversary and transfer learning techniques to mitigate the lack of sentence pairs in LRL corpora. We propose a new Low resource, Adversarial, Cross-lingual (LAC) model for NMT. In terms of the adversary technique, LAC model consists of a generator and discriminator. The generator is a Seq2Seq model that produces the translations from source to target languages, while the discriminator measures the gap between machine and human translations. In addition, we introduce transfer learning on LAC model to help capture the features in rare resources because some languages share the same subject-verb-object grammatical structure. Rather than using the entire pretrained LAC model, we separately utilize the pretrained generator and discriminator. The pretrained discriminator exhibited better performance in all experiments. Experimental results demonstrate that the LAC model achieves higher Bilingual Evaluation Understudy (BLEU) scores and has good potential to augment LRL translations.
Highlights
Traditional Neural Machine Translation (NMT) models directly learn and fit the correspondence between source and target language pairs through deep neural networks
The aforementioned approaches require a large amount of parallel bilingual data for training. For It is laborious for LowResource Languages (LRL) to build an adequate corpus for training satisfactory models
This study proposes a novel Low resource, Adversarial, and Cross-lingual Neural Machine Translation (LAC) model for NMT
Summary
Traditional Neural Machine Translation (NMT) models directly learn and fit the correspondence between source and target language pairs through deep neural networks This approach is based on a sequence-to-sequence (Seq2Seq) architecture which is comprised of encoder and decoder networks. This study presents research on adversarial learning, which achieves a higher performance in image generation [9]. It incorporates rival losses during training and can yield more explicit images. This study proposes a novel Low resource, Adversarial, and Cross-lingual Neural Machine Translation (LAC) model for NMT. Sci. 2021, 11, 10860 gradient problem in a generator Their model successfully applied adversarial learning to an NMT and achieved better translation scores. Some works explore the risk factors in machine learning models that influence the class identification in an imbalanced dataset [23,24,25]
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
