Emergency-affected population identification and notification by using online social networks

Abstract

Natural disasters have been a major cause of huge losses for both people's life and property. There is no doubt that the importance of Emergency Warning System (EWS) has been considered more seriously than ever. Unfortunately, most EWSs do not provide acceptable service to identify people who might be affected by a certain disasters. In this project, we propose an approach to identify possibly affected users of a target disaster by using online social networks. The proposed method consists of three phases. First of all, we collect location information from social network websites, such as Twitter. Then, we propose a social network analysis algorithm to identify potential victims and communities. Finally, we conduct an experiment to test the accuracy and efficiency of the approach. Based on the result, we claim that the approach can facilitate identifying potential victims effectively based on data from social networking systems. © 2011 Springer-Verlag.