Enhancing situational awareness by mining big social media data in near-real time for disaster manage: a CyberGIS approach
USC ASPIRE-I Track-1 Program (Office of the Vice President for Research)
PI: Zhenlong Li (Co-I: Amir Karami)
Summary: This project aims to design and prototype a CyberGIS system to enhance the situational awareness for better disaster management by streaming, processing, analyzing, and mapping billions of geo-tagged tweets in near-real time. Due to the massive amounts of tweets to be handled and the time-sensitive processing requirement, this project will integrate high performance computing, distributed spatial analysis, and collaborative mapping in a geo-cyberinfrastructure environment. Three recent hurricanes (Hurricane Joaquin, 2015; Hurricane Hermine, 2016; Hurricane Matthew, 2016) are used as the study cases to evaluate our prototype system. With this system, we aim to provide near-real time assessment of 1) the spatial and temporal evolution of a hurricane by monitoring the volumes of hurricane-related tweets, and where and when these tweets are posted; 2) people’s movement patterns during a hurricane by tracking and analyzing the location of millions of Twitter users; and 3) the people’s needs and opinions by analyzing large volumes of Twitter messages with text mining, sentiment analysis, spatial analysis, and social network analysis.
Jiang Y., Li Z., Ye X. (2018) Understanding Demographic and Socioeconomic Bias of Geotagged Twitter Users at the County Level, Cartography and Geographic Information Science, 15230406.2018.1434834
Martín, Y., Li, Z., & Cutter, S. L. (2017). Leveraging Twitter to gauge evacuation compliance: spatiotemporal analysis of Hurricane Matthew. PLoS one, 12(7), e0181701.
Jiang Y., Li Z. Social network, activity space, sentiment, and evacuation: what can social media tell us?, Annals of American Association of Geographers (revision under review