Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection
This addresses data quality issues for neural machine translation, but it is incremental as it generalizes existing domain selection methods to denoising.
The paper tackles the problem of data noise in neural machine translation training by proposing an approach that uses trusted data and online data selection for denoising, showing significant effectiveness in training on severely noisy data.
Measuring domain relevance of data and identifying or selecting well-fit domain data for machine translation (MT) is a well-studied topic, but denoising is not yet. Denoising is concerned with a different type of data quality and tries to reduce the negative impact of data noise on MT training, in particular, neural MT (NMT) training. This paper generalizes methods for measuring and selecting data for domain MT and applies them to denoising NMT training. The proposed approach uses trusted data and a denoising curriculum realized by online data selection. Intrinsic and extrinsic evaluations of the approach show its significant effectiveness for NMT to train on data with severe noise.