Constrained Multi-Task Learning for Event Coreference Resolution

Jing Lu, Vincent Ng


Abstract
We propose a neural event coreference model in which event coreference is jointly trained with five tasks: trigger detection, entity coreference, anaphoricity determination, realis detection, and argument extraction. To guide the learning of this complex model, we incorporate cross-task consistency constraints into the learning process as soft constraints via designing penalty functions. In addition, we propose the novel idea of viewing entity coreference and event coreference as a single coreference task, which we believe is a step towards a unified model of coreference resolution. The resulting model achieves state-of-the-art results on the KBP 2017 event coreference dataset.
Anthology ID:
2021.naacl-main.356
Volume:
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Month:
June
Year:
2021
Address:
Online
Editors:
Kristina Toutanova, Anna Rumshisky, Luke Zettlemoyer, Dilek Hakkani-Tur, Iz Beltagy, Steven Bethard, Ryan Cotterell, Tanmoy Chakraborty, Yichao Zhou
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
4504–4514
Language:
URL:
https://aclanthology.org/2021.naacl-main.356
DOI:
10.18653/v1/2021.naacl-main.356
Bibkey:
Cite (ACL):
Jing Lu and Vincent Ng. 2021. Constrained Multi-Task Learning for Event Coreference Resolution. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 4504–4514, Online. Association for Computational Linguistics.
Cite (Informal):
Constrained Multi-Task Learning for Event Coreference Resolution (Lu & Ng, NAACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.naacl-main.356.pdf
Video:
 https://aclanthology.org/2021.naacl-main.356.mp4
Code
 samlee946/cmtl-event-coref