Link prediction in supernetwork: Risk perception of emergencies

Abstract

After an emergency incident occurs, how to identify risks, predict trends and scientifically cope before the crisis erupts is the basic starting point of this study. In this study, a supernetwork model of the risk perception in emergencies is innovatively constructed from the perspective of the governance of risks. This supernetwork model includes three subnetworks: the similar relationship subnetwork that is composed of newly occurring emergencies, the chain relationship subnetwork that is composed of historical emergencies and the co-occurrence relationship subnetwork that is composed of the risk elements for emergencies. Afterwards, the feature similarity algorithm is applied to quantify the relations between newly occurring emergencies and historical emergencies, and then, the link prediction algorithm is applied to predict the risk elements that may be derived from the newly occurring emergencies. This will be beneficial to enhancing the scientific accuracy of decision-making by managers when coping with emergencies risks.