CrowdBIG: crowd-based system for information gathering from the earthquake environment

Natural disasters have always threatened the lives of humans and other creatures. One of the significant challenges for quickly responding to an earthquake is the need for precise and comprehensive information. Given that part of the environmental infrastructure is destroyed, quickly acquiring the required information is a serious challenge. Due to the ubiquity of smartphones, which have sensing, processing, and communication capabilities, this paper proposes CrowdBIG, a crowdsourcing-based architecture for information acquisition from the disaster environment. CrowdBIG architecture consists of four layers: sensing, fog, cloud, and application. Given that the reliability of crowdsourcing systems is dependent on the quality of user data, detecting malicious users, as well as scoring, and selecting useful users are of great importance. The CrowdBIG system is equipped with a reputation management component, which contains two sub-components: malicious user detection and user scoring. To evaluate the CrowdBIG system, first, we validate the information acquisition and dissemination workflow of the system using a scenario-based method. We then simulate the disaster environment through several well-known scenarios. The results show that CrowdBIG can detect malicious users appropriately. The CrowdBIG system can also score non-malicious users reasonably based on their usefulness and information completeness rates. The simulation results reveal that the reliability of the CrowdBIG system is 92%. Finally, the usability evaluation survey shows that more than 80% of the participants rated the usability of the proposed information-gathering tool as good or excellent.