Privacy Preserving Population Distribution Estimation without Trusted Third Parties

2018 Fiscal Year Final Research Report

• Project/Area Number: 16K16069
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Information security
• Research Institution: National Institute of Advanced Industrial Science and Technology
• Principal Investigator: Murakami Takao 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 主任研究員 (80587981)
• Project Period (FY): 2016-04-01 – 2019-03-31
• Keywords: 位置情報プライバシー / 空間統計データ / TTP / 加工メカニズム / 分布推定

Outline of Final Research Achievements

In this work, we studied privacy preserving population distribution estimation without trusted third parties, in which users obfuscates their locations by themselves and a data collector estimates population distribution statistics based on obfuscated location data. We first focused on the iterative Bayesian method, which is a state-of-the-art distribution estimation method, and proposed a method to reduce its estimation error. We showed, both theoretically and experimentally, that the estimation accuracy is improved. We then analyzed the security of the existing obfuscation mechanisms in terms of anonymity, and showed that OptSQL has promising in terms of the capability of anonymization.

Free Research Field

プライバシー保護

Academic Significance and Societal Importance of the Research Achievements

従来の空間統計データ構築技術は，サービス提供事業者が信頼できる機関（TTP: Trusted Third Party）であると仮定しているが，情報漏洩の事故が多発している近年ではこの仮定が成立しなくなってきている．従って，本研究での成果は，ユーザにプライバシーの観点で真の安心感を与えるという大きな意義を持つ．また，その結果，より多くのユーザから大規模な位置情報を収集することが可能となるため，従来よりも高精度な空間統計データの構築も可能となる．