The Next Generation of Disaster Management and Relief Planning: Immersive Analytics Based Approach

Abstract

Managing the risks of natural disasters can be enhanced by exploring social data. The need to swiftly extract meaningful information from large amounts of data generated by social network is on the rise especially to deal with natural disasters. New methods are needed to deeply support immersive social data analytics. Moreover, big data analysis seems to be able to improve accurate decisions to disaster management systems. The aim of this research is to determine critical cases and to focus on immersive sentiment analysis for big social data using Hadoop platform and machine learning technique. In one hand, we use MapReduce for the introduced data processing step. In the other hand, we apply support vector machine algorithm for the sentiment classification. We evaluate the performance of the performed classification method using the standard classification performance metrics accuracy, precision, recall, and F-measure and Microsoft Power BI as a visualization tool.