Development of a disaster information extraction system based on social network services

Abstract

The increased use of Social Network Services has led to an increase of availability of information. In addition, information found within SNS posts can be useful for victims in disaster-affected areas. However, a method for retrieving disaster-related information has not yet been created, thus making it difficult to find disaster-related information in SNS posts. Accordingly, this paper suggests a system which can provide efficient information when disasters occur. This system calculates three attributes of disasters. First, the Filtering module can search disaster related tweets. Moreover, it can extract the attributes of disasters systematically. Second, a manager can determine the attributes of a disaster manually by using the Decision module. In this module, disaster type, location and risk level attributes are determined. Once all of the attributes are determined, the Display module shows the attributes of the disaster. This system will provide citizens with more up-to-date disaster information. © 2014 SERSC.