摘要
随着互联网技术的迅猛发展,隐私保护已成为社会以及机构越来越关心的问题,数据挖掘技术的应用使得隐私泄露问题日益突出,隐私保护是目前数据发布中隐私泄露控制技术研究的热点问题之一,而K-匿名是近年来隐私保护研究的热点。本文介绍了K-匿名的基本概念,阐述了泛化与隐匿技术,研究了基于datafly的多维属性泛化K-匿名模型,并对该模型的基本原理、缺点进行分析,做出了相应的改进,在数据预处理阶段增加泛化层限制并且在准标识符属性选取时引入近似度分析,并对改进后的K-匿名进行实验,实验结果证明改进有效提高了处理后的数据精度。
With the rapid development of Internet technology,individuals and institutions are increasingly concerned about the issue of privacy protection.With the application of data mining technology,privacy leakage is becoming more and more serious.Privacy protection is one of the hot topics in the research of privacy leakage control technology in Data Publishing,and K-anonymous is the hotspot of privacy protection research in recent years.This paper introduces the basic concept of K-anonymous,expounds the generalization and hidden technology,introduces the multidimensional attribute generalization K-anonymous model based on Datafly,and the basic principle and the disadvantages of the model are analyzed.In addition,the corresponding improvement is made,the generalization layer is limited in the data preprocessing stage,and the approximation analysis is introduced in the selection of the identifier attribute,and the improved K-anonymity experiment is carried out.The experimental results show that the improved method effectively improves the accuracy of the processed data.
作者
岳思
吴伟明
谷勇浩
YUE Si;WU Wei-ming;GU Yong-hao(School of Computer Science, Beijing University of Post and Telecommunications, Beijing Key Laboratory of intelligent communication software and multimedia ,Beijing 100876, China;National Power Grid Corp information and communication branch, Beijing 100761, China)
出处
《软件》
2017年第11期12-17,共6页
Software
基金
国家自然科学基金项目资助(61173017
61370195)
工信部通信软科学项目资助(2014-R-42
2015-R-29)
国网科技项目(SGTYHT/15-JS-191)