On-the-fly Denoising for Data Augmentation in Natural Language Understanding
This addresses the issue of noisy augmented data impairing training for NLP practitioners, offering a general solution that is incremental over existing filtering methods.
The paper tackles the problem of noisy data introduced by data augmentation in natural language understanding by proposing an on-the-fly denoising technique that uses soft labels from a teacher model trained on cleaner original data, resulting in consistent performance improvements on text classification and question-answering tasks.
Data Augmentation (DA) is frequently used to provide additional training data without extra human annotation automatically. However, data augmentation may introduce noisy data that impairs training. To guarantee the quality of augmented data, existing methods either assume no noise exists in the augmented data and adopt consistency training or use simple heuristics such as training loss and diversity constraints to filter out "noisy" data. However, those filtered examples may still contain useful information, and dropping them completely causes a loss of supervision signals. In this paper, based on the assumption that the original dataset is cleaner than the augmented data, we propose an on-the-fly denoising technique for data augmentation that learns from soft augmented labels provided by an organic teacher model trained on the cleaner original data. To further prevent overfitting on noisy labels, a simple self-regularization module is applied to force the model prediction to be consistent across two distinct dropouts. Our method can be applied to general augmentation techniques and consistently improve the performance on both text classification and question-answering tasks.