Emergency Detection and Evacuation Planning Using Social Media

Abstract

Social media platforms have become an important part of our daily lives, especially in the last decade where number of users and messages communicated are tremendously increasing. They are commonly accepted as the easiest and fastest means of sharing recent news with others. Social media users are timely informed about all incidents taking place around the world. This is possible because people tend to share incidents on the spot despite severe consequences in cases like accidents where people in close relation with potential causalities may get shocked. On the other hand, instant communication on the social media may lead to positive impact and immediate benefit. Thus, utilizing social media has a great potential to handle specific situations, like earthquakes, terrorist attacks, and civil disorders, where getting organized timely and efficiently is crucial. In this work, we propose a system that uses strengths of social media to detect emergencies, inform and lead organizations and people so that the loss and damage can be minimized. The proposed model employs multiple agents in the emergency management process. It is capable of analyzing social media posts, filtering irrelevant and unnecessary content, detecting crisis and emergency situations, summarizing them by giving detailed information about the location and impact, creating an evacuation plan where and when possible.