Research And Application Of Privacy Leakage Rational Risk Assessment For Data Publishing

The basis of data application is data opening and sharing.Data publishing certainly faces to the threat of privacy leakage.The research on risk assessment of privacy leakage for data publishing is of great significance.This paper aims to quantify privacy and availability of data from the data privacy,data utility and privacy protection,and to discuss the calculation and evaluation of privacy leakage risk.Firstly,based on the principle of metric space,a privacy metric model for data publishing is established.Secondly,based on the metric model and game theory,a rational risk assessment model for privacy leakage is proposed.Finally,the assessment model applies to the government data publishing scene.We present a privacy protection scheme based on the rational risk computing and carry out the application simulation.Main works of this paper are:(1)A privacy metric model based on data publishing is proposed.By preprocessing on the data set,designing the function of privacy preference and quantifying the level of privacy protection,the calculation methods of privacy,data utility and level of privacy protection are put forward.(2)A rational risk assessment model for privacy leakage is proposed.We calculate the amount of privacy and data by using the established metric model,construct game model between the data publisher and receiver,analysis and define the payoff of the all parties,and solve the pure strategy Nash equilibrium and mixed strategy Nash equilibrium.According to the results of the game,this paper designs the strategy of the risk level of privacy leakage,constructs the risk assessment mechanism of privacy leakage,and gives the process of risk assessment.(3)The rational risk assessment model for privacy leakage is simulated.In the government data publishing scene,taking census data as an example,according to the characteristics of each attribute and the above metric model,a privacy protection and application simulation scheme based on rational risk calculation of privacy leakage is proposed.The simulation result shows that the scheme can effectively reduce the risk of data publishing.