A Question-Answer Driven Approach to Reveal Affirmative Interpretations from Verbal Negations

Md Mosharaf Hossain, Luke Holman, Anusha Kakileti, Tiffany Kao, Nathan Brito, Aaron Mathews, Eduardo Blanco


Abstract
This paper explores a question-answer driven approach to reveal affirmative interpretations from verbal negations (i.e., when a negation cue grammatically modifies a verb). We create a new corpus consisting of 4,472 verbal negations and discover that 67.1% of them convey that an event actually occurred. Annotators generate and answer 7,277 questions % converted for 4,000 for the 3,001 negations that convey an affirmative interpretation. We first cast the problem of revealing affirmative interpretations from negations as a natural language inference (NLI) classification task. Experimental results show that state-of-the-art transformers trained with existing NLI corpora are insufficient to reveal affirmative interpretations. We also observe, however, that fine-tuning brings substantial improvements. In addition to NLI classification, we also explore the more realistic task of generating affirmative interpretations directly from negations with the T5 transformer. We conclude that the generation task remains a challenge as T5 substantially underperforms humans.
Anthology ID:
2022.findings-naacl.37
Volume:
Findings of the Association for Computational Linguistics: NAACL 2022
Month:
July
Year:
2022
Address:
Seattle, United States
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
490–503
Language:
URL:
https://aclanthology.org/2022.findings-naacl.37
DOI:
10.18653/v1/2022.findings-naacl.37
Bibkey:
Cite (ACL):
Md Mosharaf Hossain, Luke Holman, Anusha Kakileti, Tiffany Kao, Nathan Brito, Aaron Mathews, and Eduardo Blanco. 2022. A Question-Answer Driven Approach to Reveal Affirmative Interpretations from Verbal Negations. In Findings of the Association for Computational Linguistics: NAACL 2022, pages 490–503, Seattle, United States. Association for Computational Linguistics.
Cite (Informal):
A Question-Answer Driven Approach to Reveal Affirmative Interpretations from Verbal Negations (Hossain et al., Findings 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.findings-naacl.37.pdf
Software:
 2022.findings-naacl.37.software.zip
Video:
 https://aclanthology.org/2022.findings-naacl.37.mp4
Code
 mosharafhossain/afin
Data
GLUEMultiNLIQA-SRLSNLI