Research And Implementation Of Protecting Data Publishing Method For Free Space Based On Improved Differential Privacy

With the progress and development of science and technology,the era of big data has come.People get the convenience from the analysis and application of big data.At the same time,the new privacy threat brought by it comes with it.Among them,location privacy,because of its strong relevance with the user's behavior and current life trajectory,is becoming more and more popular in academia.Many scholars have proposed algorithms to protect location privacy.However,in the era of big data,attackers are more and more likely to have great background knowledge.Most privacy protection algorithms based on anonymity mechanism are unable to cope with the attack of background knowledge.Location privacy security is facing a huge threat.Differential privacy protection is a strong privacy protection algorithm against background knowledge attack.At present,there are two kinds of index structures in the research of differential privacy protection for free space dataset,namely tree index structure and grid index structure.Tree index structure and grid index structure have their unique advantages.However,the two index structures have shortcomings in balancing query efficiency and query accuracy.However,big data analysis and application have urgent needs for privacy,query efficiency and query accuracy.In addition,there is a waste of index space in grid index structure under the condition of uneven distribution.Based on this,this thesis proposes a hybrid index structure,Grid Index Based Quadtree(GIBQT),which combines the advantages of tree index and grid index and considers the distribution characteristics of free space data.In addition,to solve the problem of space waste in grid index partition under the condition of uneven data distribution,an optimized grid index scheme named Merge Grid Index(MGI)is proposed.Through the simulation experiments of three sets of data sets with different distribution characteristics,it is proved that GIBQT index has a unique advantage in the balance of query efficiency and query accuracy,and also verifies the effectiveness of MGI scheme in solving the problem of space waste caused by sparse grid.The main contributions of this thesis are as follows:(1)The GIBQT index in this thesis considers both query accuracy and query efficiency.(2)The MGI scheme can effectively solve the problem of grid index space waste while ensuring the query accuracy under the condition of uneven distribution.(3)A software system is designed and implemented to generate a release data set that meets the requirements of GIBQT.The system allows users to generate a release data set that meets the differential privacy protection according to their own needs by adjusting parameters.