Label Definitions Improve Semantic Role Labeling

Li Zhang, Ishan Jindal, Yunyao Li


Abstract
Argument classification is at the core of Semantic Role Labeling. Given a sentence and the predicate, a semantic role label is assigned to each argument of the predicate. While semantic roles come with meaningful definitions, existing work has treated them as symbolic. Learning symbolic labels usually requires ample training data, which is frequently unavailable due to the cost of annotation. We instead propose to retrieve and leverage the definitions of these labels from the annotation guidelines. For example, the verb predicate “work” has arguments defined as “worker”, “job”, “employer”, etc. Our model achieves state-of-the-art performance on the CoNLL09 dataset injected with label definitions given the predicate senses. The performance improvement is even more pronounced in low-resource settings when training data is scarce.
Anthology ID:
2022.naacl-main.411
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:
5613–5620
Language:
URL:
https://aclanthology.org/2022.naacl-main.411
DOI:
10.18653/v1/2022.naacl-main.411
Bibkey:
Cite (ACL):
Li Zhang, Ishan Jindal, and Yunyao Li. 2022. Label Definitions Improve Semantic Role Labeling. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 5613–5620, Seattle, United States. Association for Computational Linguistics.
Cite (Informal):
Label Definitions Improve Semantic Role Labeling (Zhang et al., NAACL 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.naacl-main.411.pdf
Code
 system-t/labelawaresrl
Data
FrameNet