Exploiting Dynamic Oracles to Train Projective Dependency Parsers on Non-Projective Trees

Lauriane Aufrant, Guillaume Wisniewski, François Yvon


Abstract
Because the most common transition systems are projective, training a transition-based dependency parser often implies to either ignore or rewrite the non-projective training examples, which has an adverse impact on accuracy. In this work, we propose a simple modification of dynamic oracles, which enables the use of non-projective data when training projective parsers. Evaluation on 73 treebanks shows that our method achieves significant gains (+2 to +7 UAS for the most non-projective languages) and consistently outperforms traditional projectivization and pseudo-projectivization approaches.
Anthology ID:
N18-2066
Volume:
Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers)
Month:
June
Year:
2018
Address:
New Orleans, Louisiana
Editors:
Marilyn Walker, Heng Ji, Amanda Stent
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
413–419
Language:
URL:
https://aclanthology.org/N18-2066
DOI:
10.18653/v1/N18-2066
Bibkey:
Cite (ACL):
Lauriane Aufrant, Guillaume Wisniewski, and François Yvon. 2018. Exploiting Dynamic Oracles to Train Projective Dependency Parsers on Non-Projective Trees. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers), pages 413–419, New Orleans, Louisiana. Association for Computational Linguistics.
Cite (Informal):
Exploiting Dynamic Oracles to Train Projective Dependency Parsers on Non-Projective Trees (Aufrant et al., NAACL 2018)
Copy Citation:
PDF:
https://aclanthology.org/N18-2066.pdf