Research Of The Privacy Preservation Based On Continuous Data Publication

Along with the rapid development of Internet,big data era has arrived,the data information dissemination has become a very common phenomenon.A large number of continuous data release will inevitably leak personal privacy information,continuous release of privacy-preserving has gradually taken by the attention of scientific and technological researchers.The post-release data have a high degree of privacy-preserving,lower information loss and better availability,which are important targets for data release privacy-preserving research.At present,the continuous data release privacy-preserving research is still in the initial stage,the study of more effective data release privacy-preserving algorithm is imminent.Firstly,the advantages and disadvantages of LDMICA algorithm are analyzed,and a LDICA algorithm is proposed,which is the static data set updating privacy-preserving algorithm.The LDICA algorithm uses the clustering idea of LDMICA algorithm which using the variance to calculate the weight of each attribute,and then calculating integrated value of each record.The equivalence class is divided by the sort of integrated values and the remaining records are reasonably added to the equivalence class so that each equivalence class of the published data satisfies l-diversity and has a similar record set.The LDICA algorithm is tested experimentally.It does not modify attribute values,dividing the equivalent class after partitioning and using the lossy connection to achieve privacy-preserving,there is no information loss.When the diversity parameter l takes 1/2 of the total number of sensitive attribute values,it has the optimal computational performance overhead.Then,the LDACA algorithm is proposed by combining the clustering idea of static data publication and permutation anonymization of improving Anatomy technology,which is the dynamic data set updating privacy-preserving algorithm on continuous data publication.The equivalence class is divided by the sort of integrated values and the LDACA algorithm implements a fully updated of the data.The post-release data are also updated accordingly,and the pre-release data after updated keep the same signature.In order to implement a fully updated of the data,deletion module and modification module,counterfeit statistics and incremental table module are processed step by step.When incremental table module is processed,the LDICA algorithm which each new equivalent class of the published data satisfies l-diversity is used,besides,the original equivalent classes keep the same signature.Besides,the LDACA algorithm is tested experimentally,which reduces the information loss and can effectively prevent the connection attack.Privacy leak rate of the LDACA algorithm is 20%lower than that of the M-distinct algorithm,which is much lower than 1/l.Algorithm that execution time is less than 5s with better performance can effectively play the role of privacy-preserving.