Reversing Gradients in Adversarial Domain Adaptation for Question Deduplication and Textual Entailment Tasks

Anush Kamath, Sparsh Gupta, Vitor Carvalho


Abstract
Adversarial domain adaptation has been recently proposed as an effective technique for textual matching tasks, such as question deduplication. Here we investigate the use of gradient reversal on adversarial domain adaptation to explicitly learn both shared and unshared (domain specific) representations between two textual domains. In doing so, gradient reversal learns features that explicitly compensate for domain mismatch, while still distilling domain specific knowledge that can improve target domain accuracy. We evaluate reversing gradients for adversarial adaptation on multiple domains, and demonstrate that it significantly outperforms other methods on question deduplication as well as on recognizing textual entailment (RTE) tasks, achieving up to 7% absolute boost in base model accuracy on some datasets.
Anthology ID:
P19-1556
Volume:
Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2019
Address:
Florence, Italy
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5545–5550
Language:
URL:
https://aclanthology.org/P19-1556
DOI:
10.18653/v1/P19-1556
Bibkey:
Cite (ACL):
Anush Kamath, Sparsh Gupta, and Vitor Carvalho. 2019. Reversing Gradients in Adversarial Domain Adaptation for Question Deduplication and Textual Entailment Tasks. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 5545–5550, Florence, Italy. Association for Computational Linguistics.
Cite (Informal):
Reversing Gradients in Adversarial Domain Adaptation for Question Deduplication and Textual Entailment Tasks (Kamath et al., ACL 2019)
Copy Citation:
PDF:
https://aclanthology.org/P19-1556.pdf