arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.04013cs.CRcs.CYcs.DScs.LGstat.ML
收藏

属性隐私:框架与机制

Attribute Privacy: Framework and Mechanisms

Georgia Institute of Technology(佐治亚理工学院);Microsoft Research(微软研究院);The University of Melbourne(墨尔本大学);Columbia University(哥伦比亚大学)

Wanrong Zhang, Olga Ohrimenko, Rachel Cummings

摘要:

由于许多机器学习模型都是在机密和潜在敏感数据上训练的,因此确保训练数据的隐私性越来越受到关注。在大型数据集的分析过程中,保护个人隐私的方法受到了很大的关注。然而,在许多设置中,数据集的全局属性也可能是敏感的(例如,医院的死亡率,而不是数据集中特定患者的存在)。在这项工作中,我们从个人隐私出发,开始研究属性隐私,其中数据所有者担心在分析过程中泄露整个数据集的敏感属性。我们提出了定义来捕获\emph{attribute privacy}在两个相关的情况下,全局属性可能需要保护:(1)一个特定的数据集的属性和(2)数据集采样的基础分布的参数。我们还提供了两个有效的机制和一个低效的机制,满足这些设置的属性隐私。我们的研究结果基于河豚框架的一种新的用途,以解释数据中属性之间的相关性,从而解决了\cite{kifer2014pufferfish}留下的“开发河豚实例化和一般聚合秘密算法的挑战性问题”。

英文摘要:

Ensuring the privacy of training data is a growing concern since many machine learning models are trained on confidential and potentially sensitive data. Much attention has been devoted to methods for protecting individual privacy during analyses of large datasets. However in many settings, global properties of the dataset may also be sensitive (e.g., mortality rate in a hospital rather than presence of a particular patient in the dataset). In this work, we depart from individual privacy to initiate the study of attribute privacy, where a data owner is concerned about revealing sensitive properties of a whole dataset during analysis. We propose definitions to capture \emph{attribute privacy} in two relevant cases where global attributes may need to be protected: (1) properties of a specific dataset and (2) parameters of the underlying distribution from which dataset is sampled. We also provide two efficient mechanisms and one inefficient mechanism that satisfy attribute privacy for these settings. We base our results on a novel use of the Pufferfish framework to account for correlations across attributes in the data, thus addressing "the challenging problem of developing Pufferfish instantiations and algorithms for general aggregate secrets" that was left open by \cite{kifer2014pufferfish}.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑