Citation

Karimiziarani M, Moradkhani H. Int. J. Disaster Risk Reduct. 2023; 95: e103865.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.ijdrr.2023.103865

PMID

unavailable

Abstract

During natural disasters, people use social media platforms to share their opinions and general information about the event. Here, we investigate the public response to large and destructive hurricane Ian in late September 2022 by examining the textual content of tweets shared on Twitter across the contiguous United States (CONUS). We mined and processed over twenty million tweets for discovering the main topics of discussion and relationship between them, and classifying tweets into humanitarian topics and categories to help disaster management with thorough sentiment analysis. We employed a variety of algorithms in Artificial Intelligence for Natural Language Processing (NLP) including sentiment analysis, topic modeling, and text classification to assimilate the information content in massive Twitter data. The findings of this study provide insights on how people utilize social media to learn and disseminate information about hurricane events, which accordingly aid emergency responders and disaster managers in mitigating the negative consequences of such catastrophes and improving community preparedness.

Language: en

Keywords

Disaster management; Disaster response; Natural language processing; Sentiment analysis; Social media analytics; Twitter