分布式决策树挖掘的隐私保护研究

Research of Privacy-preserving in Distributed Decision-tree Mining

数据挖掘中的隐私保护是试图在不精确访问原始数据值的前提下,挖掘出准确的模式与规则。围绕分布式决策树挖掘的隐私保护问题展开研究,提出一种基于同态加密技术的决策树挖掘算法,使各参与方在不共享其隐私信息的前提下达到集中式挖掘的效果。理论分析和实验结果表明,该算法具有很好的隐私性、准确性和适用性。

Privacy-preserving data mining is discovering accurate patterns and rules without precise access to the original data. This paper focused on privacy-preserving research in the situation of distributed decision-tree mining, and presented a decision-tree mining algorithm based on homomorphic encryption technology, which can get accurate mining effect in the premise of no sharing of private information among mining participators. Theoretical analysis and experiment results show that this algorithm can provide good capability of privacy-preserving, accuracy and efficiency.