Summary of High-quality Data Augmentation For Low-resource Nmt: Combining a Translation Memory, a Gan Generator, and Filtering, by Hengjie Liu and Ruibo Hou and Yves Lepage
High-Quality Data Augmentation for Low-Resource NMT: Combining a Translation Memory, a GAN Generator, and Filtering
by Hengjie Liu, Ruibo Hou, Yves Lepage
First submitted to arxiv on: 22 Aug 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed approach utilizes a monolingual corpus on the source side to assist Neural Machine Translation (NMT) in low-resource settings. This is achieved by employing a Generative Adversarial Network (GAN), which augments training data for the discriminator while mitigating interference from low-quality synthetic translations with the generator. The integration of Translation Memory (TM) with NMT increases available data for the generator. A novel procedure filters synthetic sentence pairs during augmentation, ensuring high-quality data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper proposes a new way to help Neural Machine Translation (NMT) when there’s not much training data. It uses a big collection of text in the same language on the source side to improve NMT. This is done by using a special kind of computer program called a Generative Adversarial Network (GAN). The GAN helps make sure that the fake translations aren’t ruining the real ones. The paper also combines this new approach with something called Translation Memory, which gives the generator more data to work with. Finally, it has a way to filter out bad synthetic sentence pairs, so the data stays good. |
Keywords
» Artificial intelligence » Gan » Generative adversarial network » Translation
