A novel risk assessment model for privacy security in Internet of Things

Abstract

Aiming at the issues of privacy security in Internet of Things (IoT) applications, we propose an effective risk assessment model to handle probabilistic causality of evaluation factors and derive weights of influence-relation of propagation paths. The model undertakes probabilistic inference and generates values of risk probability for assets and propagation paths by using Bayesian causal relation-network and prior probability. According to Bayesian network (BN) structure, the risk analysts can easily find out relevant risk propagation paths and calculate weight values of each path by using decision-making trial and evaluation laboratory (DEMATEL). This model is applied to determine the risk level of assets and each risk propagation path as well as implement countermeasure of recommendation in accordance with evaluation results. The simulation analysis shows that this model efficiently revises recommendation of countermeasures for decision-makers and mitigates risk to an acceptable range, in addition, it provides the theoretical basis for decision-making of privacy security risk assessment (PSRA) for further development in IoT area.