Research And Implementation Of Data Anonymized Privacy Protection Method

Privacy protection means that the information about individual or collectivity should be protected.Privacy preserving in data release and mining is becoming a hot area of research in the field of information security currently.K-anonymity,l-diversity and t-closeness are the major anonymous privacy protection technology,they can reduce the risk of leakage about users' privacy information effectively.It can also guarantee the authenticity and availability of data at the same time.In addition,since differential privacy can be proved strictly by the mathematical method and solves the problem of traditional privacy protection methods that they need assumptions of background and have no ideal with measuring the level of privacy protection,it has quickly become a focus in the field of privacy protection as well as anonymous methods.But current privacy protection methods' data security and privacy protection effect still exists much space to improve.This paper proposes a new method of privacy protection based on Laplace mechanism which belongs to differential privacy in order to solve the problem about background knowledge attack.The new method designs a monotonic generalization hierarchy structure of input dataset,optimizes the data representation and gets the local-optimal transformation by comparing the generalization hierarchy.Then,considering the shortcomings that the generalization mechanism of data lead to a lot of information loss in the t-closeness privacy protection method,this paper proposed a fuzzy t-closeness privacy protection method based on fuzzy theory.The method puts forward the definitions of data fuzzification and designs the specific process of blurring the data which is used to replace the process of generalization.According to the result of fuzzy clustering,the process of blurring the data divides the fuzzy equivalence-class which meets the requirements of t-closeness privacy protection method,and then calculate the average of each element by limiting the membership degree in order to get the succedaneous point.At the same time of the above process,this paper also puts forward a construction method about fuzzy tree of semantic structure for classification attribute.Finally,this paper designs a data anonymous privacy protection module about enterprise resource planning system in order to execute the methods proposed by this paper and combine those methods with practical application.The purpose of this system is batch processing data.By designing and packaging a variety of privacy protection model,it solves lots of problems such as the lack of specification and serious resource consumption.