Research And Implementation Of Differential Privacy Techniques For Key-Value Data

The frequent occurrence of privacy data leakage incidents and the successive release of data security laws and regulations have made people pay more and more attention to personal data privacy and security issues.big challenge.Differential privacy,as the most popular privacy protection technology at present,has attracted extensive attention of scholars in the field of privacy protection because it is not affected by the background knowledge of attackers.Therefore,this paper will carry out research and analysis on differential privacy protection methods in data collection and distribution.Firstly,this paper briefly introduces the development and background of differential privacy technology,and sorts out relevant domestic and foreign research literature on centralized and local differential privacy technology in recent years.The research direction is summarized,the research points of this paper are introduced,and the relevant principles of differential privacy technology and its commonly used privacy protection mechanisms are briefly described.Secondly,this paper designs a differential privacy hierarchical collection method RHC-KV,and elaborates the processing flow of the hierarchical mechanism and perturbation mechanism contained in the method.The grading mechanism can achieve differentiated disturbance processing for data with different sensory attributes and data from different data collectors by grading the security and trustworthiness of the data during the collection process.The perturbation mechanism perturbs the original data through the privacy protection parameters determined in the grading mechanism,which not only solves the problem of low data utility and security caused by data consistency processing but also avoids data privacy leakage caused by multiple data collection risks of.Finally,the problem of low utility caused by dividing the privacy budget in the traditional method is solved by means of key-value correlation perturbation.Then the security of the method is theoretically analyzed and proved,and the effectiveness of the method is verified by simulation experiments.Finally,this paper designs the differential privacy hierarchical publishing method LHP-KV,and elaborates the key processing flow of the method.Among them,the grading mechanism follows the idea of Chapter 3 RHC-KV to solve the problem of low utility caused by data consistency processing,and at the same time avoids the risk of data privacy leakage caused by multiple queries.In the perturbation mechanism,in view of the shortcomings of the existing mean query methods,a mean query method that satisfies differential privacy is proposed.Then the security of the method is theoretically analyzed and proved,and the effectiveness of the method is verified by simulation experiments.