Privacy Protection Of Multi-party Data Integration Based On K-anonymous

At present,with the development of science and technology,all human actions are recorded by the data,such as user's shopping history and ordering data?medical data?location information,etc.Data has become a huge wealth.In order to provide better information service and make more accurate judgment,data fusion of enterprises has become a new trend of development.However,it is easy to cause the leakage of user's privacy information in the process of data infusion.In less serious case it reduces user's experience and the enterprise prestige.In serious case it threats the property of the users,affects development direction of the enterprise in the future.Therefore,research on the security of the data fusion is becoming significant.This paper summarizes the existing methods used in the data security fusion model,then analyzes the deficiency existing in the current model,and proposes two improvements about vertical data secure fusion problem: 1.Combining with the existing scheme about the secure data fusion of the two party,this paper proposes a method of multi-source data fusion method based on k-anonymous privacy model.2.Without considering sensitive value of the privacy protection problem in the data fusion process,a method of multi-source data fusion is proposed to deal with sensitive values.In view of serious time resource cost problem of existing ASHP agreement,this paper produce attribute classification trees beforehand,to decrease the number of invalid node data update,to shorten execution time of the data fusion agreement.At the same time,this paper proposed a multi-source data fusion model based on k-anonymous using the mentioned idea in ASHP,and this paper put forward a multi-source data fusion method based on the k-anonymous on the basis of ASHP protocol to realize secure fusion of multi-party.Experiments show that the method has less time than ASHP agreement in the realization of the data fusion of two party,and this method can effectively solve the question of multi-source data secure fusion.In view of the sensitive values of classified attributes,sensitive attribute values are divided into groups.This paper set different frequency constraints for different groups,and put forward the(p,A)-sensitive-k anonymous model.In order to separating sensitive value of continuous sensitive attributes and setting equivalent series constraints,the l-sensitive-k anonymous model is presented.Besides,combining TDS model,a multi-source data fusion model for sensitive value is put forward.This paper also construct a multi-source data fusion protocol.The related experiments results show that this method can not only realize the security fusion of multi-source data,but also protect the privacy of sensitive values.