| Under the background of rapid development of computer network technology,the Internet has collected a great deal of information from users in order to bring more and more convenience to people.Due to the openness of the Internet,the sharing of user information has become more and more easily,and the risk of leakage of sensitive personal information of users has greatly increased.In order to conduct data mining,data research and scientific research,in some cases,data owners need to publish data sets containing personal information of users on the Internet.In this case,personal privacy information of users is more likely leaked through the information sharing.Therefore,before these data issuing agencies publishing data,they need to hide the user's sensitive data from the original dataset in a certain way.The privacy protection in data publishing has also become the main research content in the privacy field.The main goal is to increase the availability of data in the published data while ensuring the privacy of user information is not leaked,to achieve efficient and secure information sharing.In the actual application of data publishing,it is often the case that the dataset contains multi sensitive privacy attributes and personalized protection for sensitive information.Therefore,multi sensitive attribute privacy protection and personalized publishing are hot topics in current data publishing research.This paper focuses on the existing multi sensitive bucket data technology and L-sensitive clustering method.A new grouping algorithm(BES)based on the Bigraph-similar Edges Selection is proposed,and the effectiveness of BES algorithm is proved by experiments.In the actual application of data publishing,it is often the case that the dataset contains multi sensitive privacy attributes and personalized protection for sensitive information.Therefore,multi sensitive attribute privacy protection and personalized publishing are hot topics in current data publishing research.This paper focuses on the existing multi sensitive bucket data technology and L-sensitive clustering method.A new grouping algorithm based on the Bigraph-similar Edges Selection(BES)is proposed,and the effectiveness of BES algorithm is proved by experiments.In this paper,we continue to analyze the existing weighted multi-dimensional bucket grouping algorithm(WMBF),the minimum selection priority grouping algorithm,the fully(a,k)-anonymity model,and so on,may to divide the tuples with high sensitivity to the same equivalence class,this situation will cause privacy attribute value tilt,vulnerable to homogeneous attacks.And the sensitivity of sensitive attribute values is only considered when making personalized programs,instead of considering the sensitivity of sensitive attributes themselves.Therefore,considering the sensitivity of sensitive value and sensitive attribute comprehensively,In order to avoid the tilt of the same group privacy attribute values in publishing data,we propose a(L,a)-diversity based on the L-diversity.Weighted Bigraph-similar Edges Selection algorithm(WBES)and its improved algorithm L-Split Weight Edge Selection grouping algorithm(L-SWES)are proposed for the implementation of the model.Experimental results show that the algorithm proposed in this paper can effectively avoid homogeneity attacks and get better data release effect.And the grouping algorithm is not affected by the dimension of sensitive attributes in execution time and maintains relatively good algorithm efficiency. |
