Disaster Events Detection using Twitter Data

Abstract

Twitter is a microblogging service that allows its user to share short messages called tweets with each other. All the tweets are visible on a public timeline. These tweets have the valuable geospatial component and particularly time critical events. In this paper, our interest is in the rapid detection of disaster events such as tsunami, tornadoes, forest fires, and earthquakes. We describe the detection system of disaster events and show the way to detect a target event from Twitter data. This research examines the three disasters during the same time period and compares Twitter activity and Internet news on Google. A significant result from this research is that emergency detection could begin using microblogging service.