TreeMix: Compositional Constituency-based Data Augmentation for Natural Language Understanding

Le Zhang, Zichao Yang, Diyi Yang


Abstract
Data augmentation is an effective approach to tackle over-fitting. Many previous works have proposed different data augmentations strategies for NLP, such as noise injection, word replacement, back-translation etc. Though effective, they missed one important characteristic of language–compositionality, meaning of a complex expression is built from its sub-parts. Motivated by this, we propose a compositional data augmentation approach for natural language understanding called TreeMix. Specifically, TreeMix leverages constituency parsing tree to decompose sentences into constituent sub-structures and the Mixup data augmentation technique to recombine them to generate new sentences. Compared with previous approaches, TreeMix introduces greater diversity to the samples generated and encourages models to learn compositionality of NLP data. Extensive experiments on text classification and SCAN demonstrate that TreeMix outperforms current state-of-the-art data augmentation methods.
Anthology ID:
2022.naacl-main.385
Volume:
Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Month:
July
Year:
2022
Address:
Seattle, United States
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5243–5258
Language:
URL:
https://aclanthology.org/2022.naacl-main.385
DOI:
10.18653/v1/2022.naacl-main.385
Bibkey:
Cite (ACL):
Le Zhang, Zichao Yang, and Diyi Yang. 2022. TreeMix: Compositional Constituency-based Data Augmentation for Natural Language Understanding. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 5243–5258, Seattle, United States. Association for Computational Linguistics.
Cite (Informal):
TreeMix: Compositional Constituency-based Data Augmentation for Natural Language Understanding (Zhang et al., NAACL 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.naacl-main.385.pdf
Code
 magiccircuit/treemix
Data
AG NewsGLUEQNLISCAN