Training on Synthetic Noise Improves Robustness to Natural Noise in Machine Translation
This addresses robustness issues in machine translation for noisy text, but it is incremental as it builds on existing methods like BPE and character-level encoders.
The paper tackles the problem of machine translation robustness to character-level variations like typos by training on synthetic noise, resulting in improved performance on natural noise without degrading clean-text translation.
We consider the problem of making machine translation more robust to character-level variation at the source side, such as typos. Existing methods achieve greater coverage by applying subword models such as byte-pair encoding (BPE) and character-level encoders, but these methods are highly sensitive to spelling mistakes. We show how training on a mild amount of random synthetic noise can dramatically improve robustness to these variations, without diminishing performance on clean text. We focus on translation performance on natural noise, as captured by frequent corrections in Wikipedia edit logs, and show that robustness to such noise can be achieved using a balanced diet of simple synthetic noises at training time, without access to the natural noise data or distribution.