Unsupervised Domain Adaptation Method with Semantic-Structural Alignment for Dependency Parsing

Boda Lin, Mingzheng Li, Si Li, Yong Luo


Abstract
Unsupervised cross-domain dependency parsing is to accomplish domain adaptation for dependency parsing without using labeled data in target domain. Existing methods are often of the pseudo-annotation type, which generates data through self-annotation of the base model and performing iterative training. However, these methods fail to consider the change of model structure for domain adaptation. In addition, the structural information contained in the text cannot be fully exploited. To remedy these drawbacks, we propose a Semantics-Structure Adaptative Dependency Parser (SSADP), which accomplishes unsupervised cross-domain dependency parsing without relying on pseudo-annotation or data selection. In particular, we design two feature extractors to extract semantic and structural features respectively. For each type of features, a corresponding feature adaptation method is utilized to achieve domain adaptation to align the domain distribution, which effectively enhances the unsupervised cross-domain transfer capability of the model. We validate the effectiveness of our model by conducting experiments on the CODT1 and CTB9 respectively, and the results demonstrate that our model can achieve consistent performance improvement. Besides, we verify the structure transfer ability of the proposed model by introducing Weisfeiler-Lehman Test.
Anthology ID:
2021.findings-emnlp.186
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2021
Month:
November
Year:
2021
Address:
Punta Cana, Dominican Republic
Editors:
Marie-Francine Moens, Xuanjing Huang, Lucia Specia, Scott Wen-tau Yih
Venue:
Findings
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
2158–2167
Language:
URL:
https://aclanthology.org/2021.findings-emnlp.186
DOI:
10.18653/v1/2021.findings-emnlp.186
Bibkey:
Cite (ACL):
Boda Lin, Mingzheng Li, Si Li, and Yong Luo. 2021. Unsupervised Domain Adaptation Method with Semantic-Structural Alignment for Dependency Parsing. In Findings of the Association for Computational Linguistics: EMNLP 2021, pages 2158–2167, Punta Cana, Dominican Republic. Association for Computational Linguistics.
Cite (Informal):
Unsupervised Domain Adaptation Method with Semantic-Structural Alignment for Dependency Parsing (Lin et al., Findings 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.findings-emnlp.186.pdf
Video:
 https://aclanthology.org/2021.findings-emnlp.186.mp4