隐私保护分类数据挖掘研究  

Privacy preserving classification data mining research

作　　者：张晓琳[1] 汤彪[1] 

机构地区：[1]内蒙古科技大学信息工程学院,内蒙古包头014010

出　　处：《内蒙古科技大学学报》2008年第4期329-333,共5页Journal of Inner Mongolia University of Science and Technology

基　　金：国家社会科学基金资助项目(07XTQ003);内蒙古自然科学基金重点资助项目

摘　　要：随着数据挖掘应用领域的扩大,隐私保护的数据挖掘技术研究变得越来越重要.作为隐私保护数据挖掘的主要类型——隐私保护的分类数据挖掘已经成为近年来数据挖掘领域的热点之一.如何对原始数据进行变换,然后在变换后的数据集上构造判定树是隐私保护分类数据挖掘研究的重点.基于随机扰动矩阵提出一种隐私保护分类挖掘算法.该方法适用于字符型、布尔类型、分类类型和数字类型的离散数据,并且在隐私信息的保护度和挖掘结果的准确度上都有很大的提高.With the extension of the data mining practical applications, the research of the privacy preserving data mining technique becomes more and more important. As the main type of the privacy protection data mining, privacy preserving classified data mining has already become one of the hot spots in the field of data mining in recent years. How to transform the primitive real data and then structure the decision tree based on the transformed data set is the key point of the privacy preserving classified data mining. A kind of privacy preserving classified mining algorithm was proposed on the basis of the random perturbation matrix. This method is suitable to the character type, the Boolean type, the classified type, and the digital type. The protective degree of private information and the accurate degree of the result of data mining were improved to a great extent.

关 键 词：数据挖掘 隐私保护 判定树 随机扰动矩阵 

分 类 号：TP279.2[自动化与计算机技术—检测技术与自动化装置]