策略感知的差分隐私算法设计
Design of Policy-Aware Differentially Private Algorithms
摘要:
研究了误差最优差分隐私算法的设计问题。将差分隐私应用于现实世界设置的最近工作已经使用了差分隐私的变体,其适当地修改了相邻数据库的概念。为这种差分隐私的变体设计误差最优算法的问题是开放的。在本文中,我们提出了一种新的转换等价的结果,可以把查询回答的问题下差分隐私与修改后的概念的邻居查询回答下标准的差分隐私,一个大类的邻居定义。我们利用Blowfish隐私框架,概括了差分隐私。Blowfish使用{\em policy graph}来实例化相邻数据库的不同概念。我们表明,在回答工作负载$\mathbf{W}$数据库$\mathbf{x}$根据河豚策略图$G$时所产生的错误是相同的,以回答转换工作负载$f_G(\mathbf{W})$数据库$g_G(\mathbf{x})$根据标准差分隐私,其中$f_G$和$g_G$是线性变换的基础上$G$。利用这一结果,我们开发的错误有效的算法释放直方图和多维范围查询不同的河豚政策。我们相信,我们开发的工具将有助于找到机制,以回答许多其他类别的查询,在其他政策图低错误。
英文摘要:
The problem of designing error optimal differentially private algorithms is well studied. Recent work applying differential privacy to real world settings have used variants of differential privacy that appropriately modify the notion of neighboring databases. The problem of designing error optimal algorithms for such variants of differential privacy is open. In this paper, we show a novel transformational equivalence result that can turn the problem of query answering under differential privacy with a modified notion of neighbors to one of query answering under standard differential privacy, for a large class of neighbor definitions. We utilize the Blowfish privacy framework that generalizes differential privacy. Blowfish uses a {\em policy graph} to instantiate different notions of neighboring databases. We show that the error incurred when answering a workload $\mathbf{W}$ on a database $\mathbf{x}$ under a Blowfish policy graph $G$ is identical to the error required to answer a transformed workload $f_G(\mathbf{W})$ on database $g_G(\mathbf{x})$ under standard differential privacy, where $f_G$ and $g_G$ are linear transformations based on $G$. Using this result, we develop error efficient algorithms for releasing histograms and multidimensional range queries under different Blowfish policies. We believe the tools we develop will be useful for finding mechanisms to answer many other classes of queries with low error under other policy graphs.
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。