摘要： 针对垂直分布的数据，给出一种基于隐私保护的朴素贝叶斯分类协议。该协议利用同态加密、门限密码及数字信封技术，实现数据垂直分布时的数据分类，并保证不向其他方泄露任何与结果有关的信息。理论分析表明，该协议在满足安全性的同时具有较低的通信与计算复杂度。

关键词: 数据挖掘, 朴素贝叶斯分类, 语义安全, 同态加密

Abstract: Aiming at the data of vertical distribution, this paper gives a Naive Bayes Classification(NBC) protocol based on privacy preservation. This protocol uses homomorphic encryption, threshold password and digital envelope technology to realize data classification when data is vertical distribution, and it can promise not to disclose any other party with the irrelevant information. Theory analysis shows that this protocol is in safety and low communication and computing complexity.

Key words: data mining, Naive Bayes Classification(NBC), semantic security, homomorphic encryption

中图分类号: 

• TP393