Abstract: A secure distributed mining algorithm PPDMA (Privacy Preserving Distributed Mining Algorithm) for association rules is proposed. The methods incorporate cryptographic techniques to minimize the information shared. By encrypting the constrained sub-tree and other information being transmitted, PPDMA can secure user’s privacy. Theory analysis shows the feasibility and effectiveness of the algorithm.

Key words: Data mining, Distributed database, Security, Privacy

摘要： 提出了分布式数据库的关联规则的安全挖掘算法PPDMA，通过应用密码学方法对站点间传送的用于挖掘全局频繁项集的被约束子树及其它信息进行加密，增加“干扰”信息，在接收站点对加密信息进行解密，达到不披露用户信息的目的，起到保护用户隐私的作用，以进行关联规则的安全挖掘。分析表明，该算法是正确可行的。

关键词: 数据挖掘, 分布式数据库, 安全, 隐私

CLC Number: 

• TP311