MANAGING HUMAN INDUCED CRISIS WITH BIG DATA INFRASTRUCTURE

Abstract

This paper presents human induced crisis management system using big data infrastructure. This approach was motivated by the already established fact that human induced crisis are characterized by velocity, variety and volume. This paper therefore employed Hadoop big data stack, web technology to design and implement a crisis management model. The resulting system comprises analytical engine, custom website and a desktop application called “Channel”. The Hadoop distributed file system was used for data storage in the analytical engine, crisis data were collected via Twitter API and web service generated by the project website using Apache Flume and Channel respectively. Apache Hive was used to analyse the collected data and the analysed result were posted back to custom website using Channel. The system was evaluated using Mean Opinion Score (MOS) to test for its applicability, usability and reliability. The perceived applicability rating of 74%, usability rating of 73% and reliability rating of 57% were obtained. The resulting system provides insight into crisis situation; promote rapid situational awareness, aid policy formulation and monitoring.