서브메뉴
검색
Making Differential Privacy Usable Through Human-Centered Tools
Making Differential Privacy Usable Through Human-Centered Tools
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211152721
- ISBN
- 9798384016533
- DDC
- 004
- 서명/저자
- Making Differential Privacy Usable Through Human-Centered Tools
- 발행사항
- [Sl] : Northwestern University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 170 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-02, Section: B.
- 주기사항
- Advisor: Hullman, Jessica.
- 학위논문주기
- Thesis (Ph.D.)--Northwestern University, 2024.
- 초록/해제
- 요약It is often useful to learn patterns about a population while protecting individuals' privacy.Differential privacy is a state-of-the-art framework for limiting how much information is revealed about individuals during analysis. Under differential privacy, statistical noise is injected into analyses to obscure individual contributions while maintaining overall patterns. The amount of noise is calibrated by a unit-less privacy loss parameter, ϵ, which controls a tradeoff between strength of privacy protection and accuracy of estimates. This tradeoff is difficult to reason about because it is probabilistic, non-linear, and inherently value-laden. However, people across the data ecosystem must be able to effectively reason about it in order for differential privacy to be broadly usable.Moreover, applying differential privacy in real-world settings introduces a host of socio technical challenges around communicating its guarantees and its use more broadly.To make differential privacy usable, we develop human-centered tools for data curators,data analysts, and data subjects to reason about differential privacy. Specifically, we present (1) an interactive visualization interface for data curators setting ϵ, (2) an interactive paradigm instantiated in an interactive visualization interface for analysts to spend ϵ efficiently during exploratory analysis, and (3) explanations of ϵ's privacy guarantees for data subjects. Furthermore, we present(4) an analysis of debates around the U.S. Census Bureau's use of differential privacy for the 2020 census to propose communication strategies that can facilitate more productive discussions and ensure smoother deployments going forward. In sum, this dissertation aims to increase the usability of differential privacy as a promising approach with potential to promote data privacy.
- 일반주제명
- Computer science
- 일반주제명
- Communication
- 키워드
- Data privacy
- 키워드
- Usability
- 기타저자
- Northwestern University Computer Science
- 기본자료저록
- Dissertations Abstracts International. 86-02B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017163537
■00520250211152721
■006m o d
■007cr#unu||||||||
■020 ▼a9798384016533
■035 ▼a(MiAaPQ)AAI31489689
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a004
■1001 ▼aNanayakkara, Priyanka.
■24510▼aMaking Differential Privacy Usable Through Human-Centered Tools
■260 ▼a[Sl]▼bNorthwestern University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a170 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-02, Section: B.
■500 ▼aAdvisor: Hullman, Jessica.
■5021 ▼aThesis (Ph.D.)--Northwestern University, 2024.
■520 ▼aIt is often useful to learn patterns about a population while protecting individuals' privacy.Differential privacy is a state-of-the-art framework for limiting how much information is revealed about individuals during analysis. Under differential privacy, statistical noise is injected into analyses to obscure individual contributions while maintaining overall patterns. The amount of noise is calibrated by a unit-less privacy loss parameter, ϵ, which controls a tradeoff between strength of privacy protection and accuracy of estimates. This tradeoff is difficult to reason about because it is probabilistic, non-linear, and inherently value-laden. However, people across the data ecosystem must be able to effectively reason about it in order for differential privacy to be broadly usable.Moreover, applying differential privacy in real-world settings introduces a host of socio technical challenges around communicating its guarantees and its use more broadly.To make differential privacy usable, we develop human-centered tools for data curators,data analysts, and data subjects to reason about differential privacy. Specifically, we present (1) an interactive visualization interface for data curators setting ϵ, (2) an interactive paradigm instantiated in an interactive visualization interface for analysts to spend ϵ efficiently during exploratory analysis, and (3) explanations of ϵ's privacy guarantees for data subjects. Furthermore, we present(4) an analysis of debates around the U.S. Census Bureau's use of differential privacy for the 2020 census to propose communication strategies that can facilitate more productive discussions and ensure smoother deployments going forward. In sum, this dissertation aims to increase the usability of differential privacy as a promising approach with potential to promote data privacy.
■590 ▼aSchool code: 0163.
■650 4▼aComputer science
■650 4▼aCommunication
■653 ▼aData privacy
■653 ▼aDifferential privacy
■653 ▼aPrivacy protection
■653 ▼aUsability
■690 ▼a0984
■690 ▼a0459
■71020▼aNorthwestern University▼bComputer Science.
■7730 ▼tDissertations Abstracts International▼g86-02B.
■790 ▼a0163
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17163537▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
