Data Augmentation for Neural Machine Translation using Generative Language Model
This addresses data scarcity for Neural Machine Translation practitioners, offering an incremental improvement over existing augmentation methods.
The paper tackles the bottleneck of scarce parallel corpora in Neural Machine Translation by using prompt-based data augmentation with ChatGPT to generate synthetic data, improving the baseline by 0.68 BLEU score without additional training costs.
Despite the rapid growth in model architecture, the scarcity of large parallel corpora remains the main bottleneck in Neural Machine Translation. Data augmentation is a technique that enhances the performance of data-hungry models by generating synthetic data instead of collecting new ones. We explore prompt-based data augmentation approaches that leverage large-scale language models such as ChatGPT. To create a synthetic parallel corpus, we compare 3 methods using different prompts. We employ two assessment metrics to measure the diversity of the generated synthetic data. This approach requires no further model training cost, which is mandatory in other augmentation methods like back-translation. The proposed method improves the unaugmented baseline by 0.68 BLEU score.