Separately Parameterizing Singleton Detection Improves End-to-end Neural Coreference Resolution

Xiyuan Zou, Yiran Li, Ian Porada, Jackie Cheung


Abstract
Current end-to-end coreference resolution models combine detection of singleton mentions and antecedent linking into a single step. In contrast, singleton detection was often treated as a separate step in the pre-neural era. In this work, we show that separately parameterizing these two sub-tasks also benefits end-to-end neural coreference systems. Specifically, we add a singleton detector to the coarse-to-fine (C2F) coreference model, and design an anaphoricity-aware span embedding and singleton detection loss. Our method significantly improves model performance on OntoNotes and four additional datasets.
Anthology ID:
2024.naacl-short.19
Volume:
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)
Month:
June
Year:
2024
Address:
Mexico City, Mexico
Editors:
Kevin Duh, Helena Gomez, Steven Bethard
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
212–219
Language:
URL:
https://aclanthology.org/2024.naacl-short.19
DOI:
10.18653/v1/2024.naacl-short.19
Bibkey:
Cite (ACL):
Xiyuan Zou, Yiran Li, Ian Porada, and Jackie Cheung. 2024. Separately Parameterizing Singleton Detection Improves End-to-end Neural Coreference Resolution. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers), pages 212–219, Mexico City, Mexico. Association for Computational Linguistics.
Cite (Informal):
Separately Parameterizing Singleton Detection Improves End-to-end Neural Coreference Resolution (Zou et al., NAACL 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2024.naacl-short.19.pdf