Substructure Substitution: Structured Data Augmentation for NLP

Abstract

We study a family of data augmentation methods, substructure substitution (SUB2), for natural language processing (NLP) tasks. SUB2 generates new examples by substituting substructures (e.g., subtrees or subsequences) with ones with the same label, which can be applied to many structured NLP tasks such as part-of-speech tagging and parsing. For more general tasks (e.g., text classification) which do not have explicitly annotated substructures, we present variations of SUB2 based on constituency parse trees, introducing structure-aware data augmentation methods to general NLP tasks. For most cases, training with the augmented dataset by SUB2 achieves better performance than training with the original training set. Further experiments show that SUB2 has more consistent performance than other investigated augmentation methods, across different tasks and sizes of the seed dataset.

References in corpus (6)

• Unsupervised Data Augmentation for Consistency Training
• Universal Dependencies v2: An Evergrowing Multilingual Treebank Collection
• Augmenting Data with Mixup for Sentence Classification: An Empirical Study
• Soft Contextual Data Augmentation for Neural Machine Translation
• Learning Visually-Grounded Semantics from Contrastive Adversarial Samples
• Improved Sentence Modeling using Suffix Bidirectional LSTM