Authors: Lei Zou*, Texas A&M University, Nina Lam, Louisiana State University, Heng Cai, Louisiana State University
Topics: Geographic Information Science and Systems, Hazards, Risks, and Disasters, Spatial Analysis & Modeling
Keywords: disaster resilience, social media, data mining, emergency management, Twitter
Session Type: Paper
Start / End Time: 2:35 PM / 4:15 PM
Room: Marriott Ballroom Salon 1, Marriott, Lobby Level
Presentation File: No File Uploaded
Social media such as Twitter is increasingly being used as an effective platform to observe human behaviors during disastrous events in real-time. However, there are significant theoretical as well as technical issues related to the use of social media in emergency management and disaster research. This research analyzed the Twitter use during Hurricanes Sandy and Harvey, which affected the eastern and southern coasts of United States in 2012 and 2017, respectively. We addressed three core questions: how to efficiently extract useful information from Twitter data? Did digital divide of Twitter use exist during different phases of emergency management? Can Twitter be used to improve disaster resilience through emergency rescue and post-disaster damage estimation? First, a local server to collect, store, process, and visualize Twitter data was implemented in a Hadoop cluster. Second, we developed a Twitter data mining framework to compute common Twitter indexes. Third, we examined whether geographical-social disparities existed in Twitter use during the two events. Fourth, we analyzed user behaviors of flood victims who requested for rescue on Twitter. Finally, we examined whether Twitter data could be used to monitor post-disaster damage. Results show that common indexes derived from Twitter data could enable comparison across regions and events. Social-geographical disparities in Twitter use existed in both events, with higher disaster-related Twitter use communities generally being communities of higher socioeconomic status. Twitter played a critical role in assisting non-governmental organizations and volunteers to identify and rescue victims in Harvey. Finally, social media data could improve post-disaster damage estimation.