Differential Privacy and Its Applicability for Official Statistics in Japan - A Comparative Study Using Small Area Data from the Japanese Population Census
Differential Privacy and Its Applicability for Official Statistics in Japan - A Comparative Study Using Small Area Data from the Japanese Population Census
复制标题
差异隐私及其在日本官方统计中的适用性 - 使用日本人口普查小区域数据的比较研究
DOI：
10.1007/978-3-030-57521-2_24
复制
发表时间：
2020
期刊：
影响因子：
56.9
通讯作者：
Masayuki Terada
中科院分区：
文献类型：
--
作者：
Shinsuke Ito;Takayuki Miura;Hiroto Akatsuka;Masayuki Terada
作为2020年美国人口普查准备工作的一部分，美国人口普查局使用差异隐私的方法来创建隐私保护的官方微观数据。今后，在日本，官方统计数据中使用差别隐私权的问题也将成为一个课题。本文就目前关于官方统计数据中使用差别隐私权的讨论进行综述。我们描述了美国人口普查局使用的差分隐私方法，开发了一种将差分隐私应用于日本人口网格数据的方法，并通过将其应用于日本小区域数据，对不同差分隐私方法进行了比较。结果表明，对于人口网格数据，非负小波方法（及其public-n变体）和美国人口普查自上而下的方法保留了原始数据的非负性和稀疏性，但每一种都会导致不同程度的误差。这为在三种备选办法中选择最适当的方法提供了一个重要标准。
As part of its preparations for the 2020 U.S. Population Census, the U.S. Census Bureau uses the methodology of differential privacy to create privacy-preserved official microdata. It is expected that the use of differential privacy for official statistical data will become a topic in Japan in the future.In this paper, we survey the current discussion on the use of differential privacy for creating official statistical data. We describe the differential privacy method used by the U.S. Census Bureau, develop a method to apply differential privacy to Japanese population grid data, and conduct a comparison between different differential privacy methods by applying them to Japanese small area data.Results demonstrate that for population grid data, both the non-negative Wavelet method (and its public-n variant) and the US census top-down method preserve the non-negativity and sparsity of the original data, but each results in a different degree of error. This provides an important criterion for choosing the most appropriate method among the three options.