Exploring twitter georeferenced data related to flood events: an initial approach

Abstract

The purpose of this research-in-progress paper is to present an initial data exploration of twitter georeferenced data related to flood events, in order to determine the potential of these type of data in flood damage assessment. Data exploration aimed at determine: (1) if the features of information associated to a flood event can be generated using Twitter, (2) What are the most frequently used hash tags related to flood events? and (3) If it is possible to detect post flood information from Twitter messages. We developed a script to gather data using the Twitter Search API. The data collected gave an idea of the main feature information, spatial distribution and better search strategies.