Graph Refinement for Coreference Resolution

Lesly Miculicich, James Henderson


Abstract
The state-of-the-art models for coreference resolution are based on independent mention pair-wise decisions. We propose a modelling approach that learns coreference at the document-level and takes global decisions. For this purpose, we model coreference links in a graph structure where the nodes are tokens in the text, and the edges represent the relationship between them. Our model predicts the graph in a non-autoregressive manner, then iteratively refines it based on previous predictions, allowing global dependencies between decisions. The experimental results show improvements over various baselines, reinforcing the hypothesis that document-level information improves conference resolution.
Anthology ID:
2022.findings-acl.215
Volume:
Findings of the Association for Computational Linguistics: ACL 2022
Month:
May
Year:
2022
Address:
Dublin, Ireland
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2732–2742
Language:
URL:
https://aclanthology.org/2022.findings-acl.215
DOI:
10.18653/v1/2022.findings-acl.215
Bibkey:
Cite (ACL):
Lesly Miculicich and James Henderson. 2022. Graph Refinement for Coreference Resolution. In Findings of the Association for Computational Linguistics: ACL 2022, pages 2732–2742, Dublin, Ireland. Association for Computational Linguistics.
Cite (Informal):
Graph Refinement for Coreference Resolution (Miculicich & Henderson, Findings 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingestion-script-update/2022.findings-acl.215.pdf
Video:
 https://preview.aclanthology.org/ingestion-script-update/2022.findings-acl.215.mp4
Data
OntoNotes 5.0