基于多方安全计算隐私保持的异常检测算法研究

Privacy-Preserving Outlier Detection Based on Secure Multi-Party Computation

本文首先介绍了保持隐私的异常检测算法和多方安全计算,接着提出了一种新的安全和计算算法。该算法有效地解决了多个站点间保持隐私的和计算的串谋问题,并应用于基于垂直划分的分布式数据库的异常检测中,给出了一种保持隐私异常检测算法,最后对该算法的安全性、计算和通讯开销做了简要的估计。

This paper first introduces the privacy-preserving outlier detection and secure multi-party computation, and then proposes a new algorithm for secure addition computation, which resolves the collusion in multi-party addition effectively. By applying it to the privacy-preserving outlier detection over horizontally partitioned databases, we present a secure algorithm for such a detection. Finally,we give the computational complication and the cost of communications.