摘要： 基于隐私保护的数据挖掘(PPDM)的目标是在保护原始数据的情况下建立挖掘模型并得到理想的分析结果。该文从PPDM的总体需求出发，基于数据隐藏，将PPDM技术分为安全多方计算技术、匿名技术和数据转换技术。从准确性、隐私性和复杂性3个方面对PPDM技术进行了评估。

关键词: 隐私保护, 数据挖掘, 分类, 评价

Abstract: The purpose of Privacy Preserving Data Mining(PPDM) is discovering accurate patterns and perfect results without precise access to the original data. The requirements of PPDM are put forward. Based on data hiding, PPDM methods are divided into three types: SMC, anonymization and data distortion, and they are evaluated under accuracy, privacy and complexity.

Key words: privacy preserving, data mining, classification, evaluation

中图分类号: 

• TP311