Detecting Perceived Emotions in Hurricane Disasters

Shrey Desai, Cornelia Caragea, Junyi Jessy Li


Abstract
Natural disasters (e.g., hurricanes) affect millions of people each year, causing widespread destruction in their wake. People have recently taken to social media websites (e.g., Twitter) to share their sentiments and feelings with the larger community. Consequently, these platforms have become instrumental in understanding and perceiving emotions at scale. In this paper, we introduce HurricaneEmo, an emotion dataset of 15,000 English tweets spanning three hurricanes: Harvey, Irma, and Maria. We present a comprehensive study of fine-grained emotions and propose classification tasks to discriminate between coarse-grained emotion groups. Our best BERT model, even after task-guided pre-training which leverages unlabeled Twitter data, achieves only 68% accuracy (averaged across all groups). HurricaneEmo serves not only as a challenging benchmark for models but also as a valuable resource for analyzing emotions in disaster-centric domains.
Anthology ID:
2020.acl-main.471
Volume:
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2020
Address:
Online
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5290–5305
Language:
URL:
https://aclanthology.org/2020.acl-main.471
DOI:
10.18653/v1/2020.acl-main.471
Bibkey:
Copy Citation:
PDF:
https://aclanthology.org/2020.acl-main.471.pdf
Video:
 http://slideslive.com/38929226