基于隐私保护的企业间安全多方计算研究

基于现行数据隐私问题日益严重,如何防止数据挖掘过程中隐私信息的泄漏,将是一个重要的研究议题。本文主要针对关联规则挖掘技术,从安全多方计算方面探讨隐私信息的保护,提出适用于挖掘水平分割数据的保护机制。方法设计采用集中式挖掘,过程中加入信息安全技术以确保参与单位的数据隐私,以求在隐私保护和知识获取间取得一个平衡。

Study of Privacy Preserving in Secure Multi-party Computation Between Enterprises

Privacy- preserving data mining, which finds out knowledge but preserves privacy, consequently is becoming an important but difficult issue. In this paper, privacy - preserving mechanisms for association rule mining were investigated. A mechanism that partitions data horizontally and transforms individual data for the co- operative mining was proposcd. A centralized mining model was used, data privacy of each participating party could be protected through encryption but valuable co- operative results can be discovered. The proposed mechanisms thus struck a balance between privacy protection and knowledge discovery.