Journal of Remote Sensing & GIS

Enhancing situational awareness by mining big social media data in near-real time for disaster management

3^rd World Congress on GIS and Remote Sensing

September 20-21, 2017 Charlotte, USA

Zhenlong Li

University of South Carolina, USA

Scientific Tracks Abstracts: J Remote Sensing & GIS

Abstract:

Situational awareness (SA) is critical for local authorities and emergency responders to identify areas in need of immediate attention during a disaster. However, the level of SA is mainly depended on the availability of real-time data sources, and the ability to collect, synthesize and map these data in real-time. Traditional data collection practices such as field surveying and remote sensing often fail to offer timely information during or right after a disaster event. Social media (e.g. Twitter), which captures micro-level, real-time information using �??human-as-sensors�?� during a disaster (e.g., hurricane), is emerging as a new data source for natural disaster management. By integrating spatial analysis, data mining, and high-performance computing, we developed a series of methods and a prototype system to rapidly collect, analyze and map billions of geotagged tweets during the rapid onset disasters, which is able to provide emergency managers and other more relevant and possibly better decision support information. Specifically, in this presentation, I will use four research examples to demonstrate how such a system could enhance disaster situational awareness: 1) streaming and managing billions of tweets for real-time query analytics; 2) rapidly exploring spatiotemporal patterns of twitter activity during a disaster; 3) leveraging geotagged tweets for near-real-time flood mapping using 2015 Hurricane Joaquin as case study and; 4) analyzing and mapping the evacuation behavior of South Carolina resident during the 2016 Hurricane Matthew.

Biography :

Zhenlong Li is an Assistant Professor in the Department of Geography at the University of South Carolina. He is the founding Director of Geoinformation and Big Data Research Lab. His research focuses on spatial high-performance/cloud computing, big data processing/mining, and geospatial cyberinfrastructure within the area of data and computational intensive GISciences, aiming to optimize spatial computing infrastructure by integrating cutting-edge computing technologies and spatial principles to support domain applications such as climate change and hazard management. He has co-edited two books, published over 20 peer reviewed journal articles, and over 15 book chapters and peer-reviewed conference papers.