现代电子技术2025，Vol.48Issue(18)：114-118,5.DOI:10.16652/j.issn.1004-373x.2025.18.018
基于属性分割的差分隐私高维数据发布方法
Method of DP high dimensional data dissemination based on attribute segmentation
摘要
Abstract
In order to address the phenomenon of decreased data availability during high-dimensional data dissemination,a method of differential privacy high-dimensional data dissemination is proposed.This method is based on attribute segmentation technology and guided by differential privacy(DP)algorithm.The high-dimensional datasets are divided based on attribute sensitivity,and the information entropy and maximum information coefficient are used to identify the sensitivity of attributes and the dependence between attributes.According to the sensitivity of the dataset,it is divided into three types:sensitive,general sensitive,and low sensitive datasets.The Bayesian network with DP is created,and the noise-adding conditions for privacy budget are generated to complete the secure sharing of high-dimensional data under differential privacy mechanism.The experimental results show that the proposed method can not only effectively control the risk of privacy leakage,but also maintain the availability of data well.关键词
属性分割/差分隐私算法/高维数据/数据发布/贝叶斯网络/隐私预算/敏感度Key words
attribute segmentation/differential privacy/high dimensional data/data dissemination/Bayesian network/privacy budget/sensitivity分类
信息技术与安全科学引用本文复制引用
张衡,胡婉如..基于属性分割的差分隐私高维数据发布方法[J].现代电子技术,2025,48(18):114-118,5.
