A Simple and Effective Approach to Automatic Post-Editing with Transfer Learning
This provides a more efficient approach for improving machine translation outputs, though it is incremental as it builds on existing transfer learning methods.
The authors tackled the problem of automatic post-editing for machine translation by fine-tuning pre-trained BERT models with parameter sharing, achieving competitive results with only 23K sentences and 3 hours of training, and state-of-the-art results when adding artificial data.
Automatic post-editing (APE) seeks to automatically refine the output of a black-box machine translation (MT) system through human post-edits. APE systems are usually trained by complementing human post-edited data with large, artificial data generated through back-translations, a time-consuming process often no easier than training an MT system from scratch. In this paper, we propose an alternative where we fine-tune pre-trained BERT models on both the encoder and decoder of an APE system, exploring several parameter sharing strategies. By only training on a dataset of 23K sentences for 3 hours on a single GPU, we obtain results that are competitive with systems that were trained on 5M artificial sentences. When we add this artificial data, our method obtains state-of-the-art results.