Tweets Analysis for Disaster Management: Preparedness, Emergency Response, Impact, and Recovery

Abstract

Social media is very important source for identifying and analyzing various disaster events. The News channels broadcast their headlines of news on twitter as short messages. In case of any disaster situation the news channel immediately broadcast the news. Social media data is categorized into the various phases of disaster management. Disaster researchers and emergency managers use social media data as reference for their analysis. To figure out the disaster management and the social media under various shifts and numerous phases and decide with the an efficient solution “real time” is required. We will be using and comparing Support Vector Machine (SVM), K-nearest neighbors (KNN) and Logistic regression, data mining algorithms for classification of tweets. The most accurate algorithm is applied on real time tweets of news channel for handling disaster situation. The analysis is beneficial for disaster management team to detect the change of state between various stages of disaster management.