Research On Privacy Protection In Governmental Data Opening

With the development of big data era,data opening has become the basis for realizing data utilization and innovative applications.Government data opening is the key to realizing the value of government data.At the same time,the demand of public?businesses and other organizations for government data opening is increasing.However,there are various privacy issues and different degrees of privacy leakage risk in the government's data opening which in turn affects open data.Therefore,how to promote the openness of government data,improve the data utilization rate,and realize the value of data under the premise of protecting the privacy of citizens has become a challenging problem in the process of data openness.Aiming at how to promote the use of government open data and meet the privacy protection needs the paper mainly completes the following research work:(1)This paper firstly analyzes the risk of privacy in government data opening studies the phenomenon and reason of low utilization rate about government open data.An open model of government data based on demand and privacy risks is proposed.Then combined with the idea of privacy protection and based on government cloud,a privacy protection data opening framework has been constructed.And the roles and functions of the data source layer,department cloud layer,government affairs center cloud layer and user layer in the framework are defined.Finally,this article analyzes the advantages of the open model used by government,and designs the privacy protection government data publishing process data open mechanism under the open models.(2)In order to solve the privacy leakage problem of multi-sensitive attribute data in the directional opening of government data,a Maximal Weighted Joint Sensitivity Bucket First algorithm(MWJSBF)is designed.This paper focuses on multi-sensitive attribute data with high risk and demand in directional opening.Firstly,the dataset is divided using the similarity between the data record quasi-identifier attribute values.For the classification attribute,based on the information entropy,we measure theweight of each attribute to the similarity.Then calculate the similarity between the data records,and use Canopy-Kmeans clustering algorithm based on Map Reduce to achieve efficient partitioning of data.Secondly,classify sensitive attributes using rough sets and attribute dependencies,and use information entropy to calculate the sensitivity of each sensitive attribute.Then each cluster sub table is divided into quasi identifier attribute table and sensitive attribute table.The multi-dimensional bucket structure is constructed based on the joint sensitivity.Calculate weighted selectivity and group until the multi-sensitive attribute l-diversity is satisfied.Finally,the identifier attribute is generalized.The experimental comparison with MBF and MMDCF algorithms shows that the method proposed in this paper can effectively reduce the data concealment rate and the degree of additional information loss,and improve the availability of open data.(3)By using game theory and Nash equilibrium theory,the parameter setting of multi-sensitive attribute privacy protection algorithm based on t-closeness is optimized.According to the application based opening process proposed in the directional opening mode and on-demand opening mode of this paper,a dynamic game model between government and data users is constructed.The utility functions of both sides of the game are analyzed and designed.Based on this,the game model is solved in equilibrium.And it is applied to the privacy protection model based on t-closeness multi-sensitive attributes,which improves the parameter setting of the original model,and achieves the balance of privacy protection and data availability on the premise of protecting the interests of data providers and data demanders.