Title: Visualization of disaster information over time by using disaster information extraction from Twitter
Authors: Yoshiro Yamamoto - Tokai Univeristy (Japan) [presenting]
Takamitsu Funayama - Tokai University (Japan)
Osamu Uchida - Tohoku University (Japan)
Abstract: Many natural disasters such as earthquakes occurred in Japan. Many cases have been reported that use SNS such as Twitter to reduce disaster in the event of a disaster. We aim to extract useful information for grasping the situation at the time of disaster from Twitter and to visualize the information to make it easier to understand. Since Twitter information is Tweeted freely by unspecified majority in the whole world, it is difficult even by limiting specific disasters to Tweet. It is necessary to extract Tweet extracted by keywords such as ``earthquake'', Tweet by the person actually affected has occurred. In visualization that can grasp the situation of the earthquake by text mining, visualization was devised aiming at grasping the change of the situation in consecutive several hours unit. We analyzed information posted on Twitter during the earthquake in Kumamoto in April 2016. We will report a case where we visualize the change of Tweet for two days from the occurrence of the earthquake.