Abstract: With the advent of the era big data,the potential value of mining big data has attracted more and more attention from academia and industry.However,at the same time,due to frequent Internet security incidents,users are increasingly concerned about the disclosure of personal privacy data,and user data security issues become one of the most important obstacles to big data analysis.With regard to the study of user data security,the existing researches more focus on access control,ciphertext retrieval and result verification.The above researches can guarantee the security of user data itself,but can not dig out the potential value of protected data.Therefore,how to protect the security and dig the potential value of the data in the meantime is one of the key issues that need to be addressed.This paper proposed an association rules mining method based on differential privacy protection.Data owners use Laplacian mechanism and exponential mechanism to protect user data during data release.Data analysis is associated with differential privacy FP-tree Rule mining.The experimental results show that the performance and accuracy of the proposed method are superior to the existing methods.

Key words: Differential privacy, Exponential mechanism, Laplace mechanism, Privacy preserving data mining