Less is More: Attention Supervision with Counterfactuals for Text Classification

Seungtaek Choi, Haeju Park, Jinyoung Yeo, Seung-won Hwang


Abstract
We aim to leverage human and machine intelligence together for attention supervision. Specifically, we show that human annotation cost can be kept reasonably low, while its quality can be enhanced by machine self-supervision. Specifically, for this goal, we explore the advantage of counterfactual reasoning, over associative reasoning typically used in attention supervision. Our empirical results show that this machine-augmented human attention supervision is more effective than existing methods requiring a higher annotation cost, in text classification tasks, including sentiment analysis and news categorization.
Anthology ID:
2020.emnlp-main.543
Volume:
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Month:
November
Year:
2020
Address:
Online
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
6695–6704
Language:
URL:
https://aclanthology.org/2020.emnlp-main.543
DOI:
10.18653/v1/2020.emnlp-main.543
Bibkey:
Cite (ACL):
Seungtaek Choi, Haeju Park, Jinyoung Yeo, and Seung-won Hwang. 2020. Less is More: Attention Supervision with Counterfactuals for Text Classification. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 6695–6704, Online. Association for Computational Linguistics.
Cite (Informal):
Less is More: Attention Supervision with Counterfactuals for Text Classification (Choi et al., EMNLP 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.emnlp-main.543.pdf
Video:
 https://slideslive.com/38939245