RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response
By Ú. Erlingsson, Aleksandra Korolova, Vasyl Pihur. 2014
2,273 citations recorded by the source corpus.
Randomized Aggregatable Privacy-Preserving Ordinal Response, or RAPPOR, is a technology for crowdsourcing statistics from end-user client software, anonymously, with strong privacy guarantees. In short, RAPPORs allow the forest of client data to be studied, without permitting the possibility of looking at individual trees. By applying randomized response in a novel manner, RAPPOR provides the mechanisms for such collection as well as for efficient, high-utility analysis of the collected data. In particular, RAPPOR permits statistics to be collected on the population of client-side strings with strong privacy guarantees for each client, and without linkability of their reports. This paper describes and motivates RAPPOR, details its differential-privacy and utility guarantees, discusses its practical deployment and properties in the face of different attack models, and, finally, gives results of its application to both synthetic and real-world data.
Subfields: Artificial Intelligence, Computer Science Applications, Safety, Robustness & Interpretability, Privacy & Security
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Percy Liang
- Philip S. Yu
- Peter Kairouz
- Mohit Bansal
- Dawn Song
- Yoshua Bengio
- Abhradeep Thakurta
- Vitaly Feldman
In Computer Science Applications
- Jayadev Acharya
- Ninghui Li
- Jonathan Ullman
- Ziteng Sun
- Tianhao Wang
- Vitaly Feldman
- Peter Kairouz
- Aleksandra Korolova
In Safety, Robustness & Interpretability
- Nicholas Carlini
- Nicolas Papernot
- Florian Tramèr
- Dawn Song
- Bo Li
- Patrick Drew McDaniel
- Neil Zhenqiang Gong
- Abhradeep Thakurta
In Privacy & Security
- Jonathan Ullman
- Abhradeep Thakurta
- Vitaly Feldman
- Peter Kairouz
- Kunal Talwar
- Borja Balle
- Badih Ghazi
- Thomas Steinke
Citations in the corpus
- Extracting Training Data from Large Language Models (2020, background): “Large companies have even used DP in production machine learning models to protect users’ sensitive information [13,64].”
- Towards the Science of Security and Privacy in Machine Learning (2016, background): “Erlingsson et al. showed that this allowed the developers of a browser to collect meaningful and privacy-preserving usage statistics from users [97].”
- Foundation Models and Fair Use (2023, background): “, Smart Compose, telemetry collection) (Miraglia, 2019; Erlingsson et al., 2014; Ding et al., 2017; Bittau et al., 2017; Cheu et al., 2019).”
- Advances and Open Problems in Federated Learning (2019, methodology): “LDP has been deployed effectively to gather statistics on popular items across large user bases by Google, Apple and Microsoft [156, 135, 136].”
- Federated Learning With Differential Privacy: Algorithms and Performance Analysis (2019, background): “The work in [27] proposed solutions to building up an LDP-compliant SGD, which powers a variety of important ML tasks.”
- Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data (2016, background): “Erlingsson et al. (2014) showed that randomized response, introduced by Warner (1965), can protect crowd-sourced data collected from software users to compute statistics about user behaviors.”
- On Distributed Differential Privacy and Counting Distinct Elements (2020, background): “Differential privacy (DP) [20, 19] has become a leading framework for private-data analysis, with several recent practical deployments [25, 39, 28, 3, 16, 1].”
- Breaking the Communication-Privacy-Accuracy Trilemma (2020, methodology): “If we further assume X are drawn from some discrete distribution p, then the problem falls into distribution estimation under local differential privacy [9], [13], [34], [36], [42], [43], [44], [45], [46] and limited communication [11], [25], [45], [46], [47], [48], [49], [50].”
- Privacy Amplification via Random Check-Ins (2020, methodology): “For this reason, DP has been adopted by companies including Google [20, 10, 18], Apple [2], Microsoft [13], and LinkedIn [31], as well as the US Census Bureau [26].”
- Private Counting from Anonymous Messages: Near-Optimal Accuracy with Vanishing Communication Overhead (2020, methodology): “In contrast, computing near-optimal parameters in the shuffled model of the noise parameters for non-infinitely divisible schemes such as Randomized Response (Warner, 1965) and RAPPOR (Erlingsson et al., 2014) requires re-running a time-expensive algorithm for each new value of n (see the supplement”
- Auditing Differentially Private Machine Learning: How Private is Private SGD? (2020, background): “Differential privacy [DMNS06] has become the de facto standard for guaranteeing privacy in machine learning and statistical analysis, and is now being deployed by many organizations including Apple [TVV + 17], Google [EPK14, BEM + 17, PSM + 18], and the US Census Bureau [HMA + 17].”
- Pure Differentially Private Summation from Anonymous Messages (2020, methodology): “It also rules out the protocol implied by shuffling RAPPOR [EPK14], and more generally any protocol obtained by the amplification via shuffling approach of [EFM + 19, BBGN19c].”
- The power of synergy in differential privacy: Combining a small curator with local randomizers (2019, background): “…of local di ff erential privacy protocols by Google, Apple, and Microsoft demonstrate that the local di ff erential privacy model is a viable approach in certain applications, without requiring trust in the companies or incurring risks from hackers and intelligence agencies [28, 55, 35, 1, 22].”
- Secure Federated Submodel Learning (2019, methodology): “In addition, with a higher heterogeneity of user data and thus a higher differentiation of submodels, the roughness and inaccuracy of conventional federated learning will be exposed more completely, which clarifies why it can work in the natural language context with a 10,000 word vocabulary consider”
- Fair decision making using privacy-protected data (2019, background): “ividuals equally? Or are some populations systematically disadvantaged as a result of privacy technology? These questions are especially important now: the adoption of differential privacy is growing [13, 14, 16, 21], and, in particular, the U.S. Census Bureau is currently exploring differentially p”