A Modest Pareto Optimisation Analysis of Dependency Parsers in 2021

Mark Anderson, Carlos Gómez-Rodríguez


Abstract
We evaluate three leading dependency parser systems from different paradigms on a small yet diverse subset of languages in terms of their accuracy-efficiency Pareto front. As we are interested in efficiency, we evaluate core parsers without pretrained language models (as these are typically huge networks and would constitute most of the compute time) or other augmentations that can be transversally applied to any of them. Biaffine parsing emerges as a well-balanced default choice, with sequence-labelling parsing being preferable if inference speed (but not training energy cost) is the priority.
Anthology ID:
2021.iwpt-1.12
Volume:
Proceedings of the 17th International Conference on Parsing Technologies and the IWPT 2021 Shared Task on Parsing into Enhanced Universal Dependencies (IWPT 2021)
Month:
August
Year:
2021
Address:
Online
Editors:
Stephan Oepen, Kenji Sagae, Reut Tsarfaty, Gosse Bouma, Djamé Seddah, Daniel Zeman
Venue:
IWPT
SIG:
SIGPARSE
Publisher:
Association for Computational Linguistics
Note:
Pages:
119–130
Language:
URL:
https://aclanthology.org/2021.iwpt-1.12
DOI:
10.18653/v1/2021.iwpt-1.12
Bibkey:
Cite (ACL):
Mark Anderson and Carlos Gómez-Rodríguez. 2021. A Modest Pareto Optimisation Analysis of Dependency Parsers in 2021. In Proceedings of the 17th International Conference on Parsing Technologies and the IWPT 2021 Shared Task on Parsing into Enhanced Universal Dependencies (IWPT 2021), pages 119–130, Online. Association for Computational Linguistics.
Cite (Informal):
A Modest Pareto Optimisation Analysis of Dependency Parsers in 2021 (Anderson & Gómez-Rodríguez, IWPT 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.iwpt-1.12.pdf
Data
Penn Treebank