Improving Generalization in Coreference Resolution via Adversarial Training

Sanjay Subramanian, Dan Roth


Abstract
In order for coreference resolution systems to be useful in practice, they must be able to generalize to new text. In this work, we demonstrate that the performance of the state-of-the-art system decreases when the names of PER and GPE named entities in the CoNLL dataset are changed to names that do not occur in the training set. We use the technique of adversarial gradient-based training to retrain the state-of-the-art system and demonstrate that the retrained system achieves higher performance on the CoNLL dataset (both with and without the change of named entities) and the GAP dataset.
Anthology ID:
S19-1021
Volume:
Proceedings of the Eighth Joint Conference on Lexical and Computational Semantics (*SEM 2019)
Month:
June
Year:
2019
Address:
Minneapolis, Minnesota
Editors:
Rada Mihalcea, Ekaterina Shutova, Lun-Wei Ku, Kilian Evang, Soujanya Poria
Venue:
*SEM
SIGs:
SIGSEM | SIGLEX
Publisher:
Association for Computational Linguistics
Note:
Pages:
192–197
Language:
URL:
https://aclanthology.org/S19-1021
DOI:
10.18653/v1/S19-1021
Bibkey:
Cite (ACL):
Sanjay Subramanian and Dan Roth. 2019. Improving Generalization in Coreference Resolution via Adversarial Training. In Proceedings of the Eighth Joint Conference on Lexical and Computational Semantics (*SEM 2019), pages 192–197, Minneapolis, Minnesota. Association for Computational Linguistics.
Cite (Informal):
Improving Generalization in Coreference Resolution via Adversarial Training (Subramanian & Roth, *SEM 2019)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-5/S19-1021.pdf