Privacy Tools for Sharing Research Data

Information technology, advances in statistical computing, and the deluge of data available through the Internet are transforming computational social science. However, a major challenge is maintaining the privacy of human subjects. This project is a broad, multidisciplinary effort to help enable the collection, analysis, and sharing of sensitive data while providing privacy for individual subjects. Bringing together computer science, social science, statistics, and law, the investigators seek to refine and develop definitions and measures of privacy and data utility, and design an array of technological, legal, and policy tools for dealing with sensitive data. In addition to contributing to research infrastructure around the world, the ideas developed in this project will benefit society more broadly as it grapples with data privacy issues in many other domains, including public health and electronic commerce.

This project will define and measure privacy in both mathematical and legal terms, and explore alternate definitions of privacy that may be more general or more practical. The project will study variants of differential privacy and develop new theoretical results for use in contexts where it is currently inappropriate or impractical. The research will provide a better understanding of the practical performance and usability of a variety of algorithms for analyzing and sharing privacy-sensitive data. The project will develop secure implementations of these algorithms and legal instruments, which will be made publicly available and used to enable wider access to privacy-sensitive data sets at the Harvard Institute for Quantitative Social Science's Dataverse Network.

This project is funded by a National Science Foundation Secure and Trustworthy Cyberspace Frontier Grant and a gift from Google. For more information, see the original proposed project description to NSF (2012).

Two major areas of research in this project are DataTags and Differential Privacy. This project has contributed to the development of the software tools DataTags, PSI, and AbcDatalog.

Senior Personnel

Salil Vadhan

Salil Vadhan

Vicky Joseph Professor of Computer Science and Applied Mathematics, SEAS, Harvard

Salil Vadhan is the lead PI of Privacy Tools for Sharing Research Data project and the Vicky Joseph Professor of Computer Science and Applied Mathematics.

Marco Gaboardi

Marco Gaboardi

Visiting Scholar, Center for Research on Computation & Society
State University of New York at Buffalo
Urs Gasser

Urs Gasser

Executive Director, Berkman Center for Internet & Society
Professor of Practice, Harvard Law School
Kobbi Nissim

Kobbi Nissim

Senior Research Fellow at Harvard University
Professor of Computer Science at Georgetown University
  •  
  • 1 of 3
  • »

Publications

C. Dwork, A Smith, T Steinke, J Ullman, and S. Vadhan. 2015. “Robust Traceability from Trace Amounts.” In IEEE Symposium on Foundations of Computer Science (FOCS 2015). Berkeley, California.Abstract

The privacy risks inherent in the release of a large number of summary statistics were illustrated by Homer et al. (PLoS Genetics, 2008), who considered the case of 1-way marginals of SNP allele frequencies obtained in a genome-wide association study: Given a large number of minor allele frequencies from a case group of individuals diagnosed with a particular disease, together with the genomic data of a single target individual and statistics from a sizable reference dataset independently drawn from the same population, an attacker can determine with high confidence whether or not the target is in the case group. In this work we describe and analyze a simple attack that succeeds even if the summary statistics are significantly distorted, whether due to measurement error or noise intentionally introduced to protect privacy. Our attack only requires that the vector of distorted summary statistics is close to the vector of true marginals in `1 norm. Moreover, the reference pool required by previous attacks can be replaced by a single sample drawn from the underlying population. The new attack, which is not specific to genomics and which handles Gaussian as well as Bernouilli data, significantly generalizes recent lower bounds on the noise needed to ensure differential privacy (Bun, Ullman, and Vadhan, STOC 2014; Steinke and Ullman, 2015), obviating the need for the attacker to control the exact distribution of the data.

Micah Altman, Alexandra Wood, David O'Brien, Salil Vadhan, and Urs Gasser. 2016. “Towards a Modern Approach to Privacy-Aware Government Data Releases.” Berkeley Technology Law Journal, 3, 30. BTLJ VersionAbstract

Transparency is a fundamental principle of democratic governance. Making government data more widely available promises to enhance organizational transparency, improve government functions, encourage civic engagement, support the evaluation of government decisions, and ensure accountability for public institutions. Furthermore, releases of government data promote growth in the private sector, guiding investment and other commercial decisions, supporting innovation in the technology sectors, and promoting economic development and competition generally. Improving access to government data also advances the state of research and scientific knowledge, changing how researchers approach their fields of study and enabling them to ask new questions and gain better insights into human behaviors. For instance, the increased availability of large-scale datasets is advancing developments in computational social science, a field that is rapidly changing the study of humans, human behavior, and human institutions, and effectively shifting the evidence base of social science. Scientists are also developing methods to mine and model new data sources and big data, and data collected from people and institutions have proven useful in unexpected ways. In the area of public health, Google Flu Trends, which provides a useful and timely supplement to conventional flu tracking methods by analyzing routine Google queries, is a widely publicized example of the unexpected uses of data. These are, of course, just a few examples of the many benefits of open data.

Jack Murtagh and Salil Vadhan. 2016. “The Complexity of Computing the Optimal Composition of Differential Privacy.” In Theory of Cryptography Conference (TCC 2016). ArXiv VersionAbstract

In the study of differential privacy, composition theorems (starting with the original paper of Dwork, McSherry, Nissim, and Smith (TCC'06)) bound the degradation of privacy when composing several differentially private algorithms. Kairouz, Oh, and Viswanath (ICML'15) showed how to compute the optimal bound for composing k arbitrary (ϵ,δ)-differentially private algorithms. We characterize the optimal composition for the more general case of k arbitrary (ϵ1,δ1),,(ϵk,δk)-differentially private algorithms where the privacy parameters may differ for each algorithm in the composition. We show that computing the optimal composition in general is #P-complete. Since computing optimal composition exactly is infeasible (unless FP=#P), we give an approximation algorithm that computes the composition to arbitrary accuracy in polynomial time. The algorithm is a modification of Dyer's dynamic programming approach to approximately counting solutions to knapsack problems (STOC'03).

Mark Bun, Kobbi Nissim, Uri Stemmer, and Salil Vadhan. 2015. “Differentially Private Release and Learning of Threshold Functions.” In 56th Annual IEEE Symposium on Foundations of Computer Science (FOCS 15). Berkeley, California. ArXiv VersionAbstract

We prove new upper and lower bounds on the sample complexity of (ϵ,δ) differentially private algorithms for releasing approximate answers to threshold functions. A threshold function cx over a totally ordered domain X evaluates to cx(y)=1 if yx, and evaluates to 0 otherwise. We give the first nontrivial lower bound for releasing thresholds with (ϵ,δ) differential privacy, showing that the task is impossible over an infinite domain X, and moreover requires sample complexity nΩ(log|X|), which grows with the size of the domain. Inspired by the techniques used to prove this lower bound, we give an algorithm for releasing thresholds with n2(1+o(1))log|X| samples. This improves the previous best upper bound of 8(1+o(1))log|X| (Beimel et al., RANDOM '13).
Our sample complexity upper and lower bounds also apply to the tasks of learning distributions with respect to Kolmogorov distance and of properly PAC learning thresholds with differential privacy. The lower bound gives the first separation between the sample complexity of properly learning a concept class with (ϵ,δ) differential privacy and learning without privacy. For properly learning thresholds in dimensions, this lower bound extends to nΩ(log|X|).
To obtain our results, we give reductions in both directions from releasing and properly learning thresholds and the simpler interior point problem. Given a database D of elements from X, the interior point problem asks for an element between the smallest and largest elements in D. We introduce new recursive constructions for bounding the sample complexity of the interior point problem, as well as further reductions and techniques for proving impossibility results for other basic problems in differential privacy.

David O'Brien, Jonathan Ullman, Micah Altman, Urs Gasser, Michael Bar-Sinai, Kobbi Nissim, Salil Vadhan, Michael Wojcik, and Alexandra Wood. 2015. “Integrating Approaches to Privacy Across the Research Lifecycle: When is Information Purely Public?.” Social Science Research Network. SSRN VersionAbstract

On September 24-25, 2013, the Privacy Tools for Sharing Research Data project at Harvard University held a workshop titled "Integrating Approaches to Privacy across the Research Data Lifecycle." Over forty leading experts in computer science, statistics, law, policy, and social science research convened to discuss the state of the art in data privacy research. The resulting conversations centered on the emerging tools and approaches from the participants’ various disciplines and how they should be integrated in the context of real-world use cases that involve the management of confidential research data.

Researchers are increasingly obtaining data from social networking websites, publicly-placed sensors, government records and other public sources. Much of this information appears public, at least to first impressions, and it is capable of being used in research for a wide variety of purposes with seemingly minimal legal restrictions. The insights about human behaviors we may gain from research that uses this data are promising. However, members of the research community are questioning the ethics of these practices, and at the heart of the matter are some difficult questions about the boundaries between public and private information. This workshop report, the second in a series, identifies selected questions and explores issues around the meaning of “public” in the context of using data about individuals for research purposes.

Micah Altman, David O’Brien, Salil Vadhan, and Alexandra Wood. 2014. “Comment to The White House Office of Science and Technology Policy (OSTP): Big Data Study, Request for Information”.Abstract
On January 23, 2014, President Barack Obama asked John Podesta to perform a comprehensive review of big data and privacy. During this review, the White House Office of Science and Technology Policy issued a request for public comment on questions related to the public policy implications of big data.
 
 
Micah Altman, David O’Brien, Salil Vadhan, and Alexandra Wood submitted a response on behalf of the Privacy Tools for Sharing Research Data project. Their comments outline a broad, comprehensive, and systematic framework for privacy analysis and provide a taxonomy of modern technological, statistical, and cryptographic approaches to preserving both data privacy and utility. They argue that an analysis of information privacy should address the scope of information covered, the sensitivity of that information, the risk that sensitive information will be disclosed, the availability of control and accountability mechanisms, and the suitability of existing data sharing models, as applied across the entire lifecyle of information use, from collection through dissemination and reuse.
 
 
With this submission, the authors discuss the inadequacy of traditional approaches to privacy protection and recommend a modern approach that considers three principles. First, the risks of informational harm are generally not a simple function of the presence or absence of specific fields, attributes, or keywords in the released set of data. Second, redaction, pseudonymization, coarsening, and hashing, are often neither an adequate nor appropriate practice, nor is releasing less information necessary more privacy protective. Third, a thoughtful analysis with expert consultation is necessary in order to evaluate the sensitivity of the data collected, to quantify the associated re-identification risks, and to design useful and safe release mechanisms.
Yiling Chen, Or Sheffet, and Salil Vadhan. 2014. “Privacy Games.” In 10th Conference on Web and Internet Economics (WINE). Beijing, China.
Alexandra Wood, David O'Brien, Micah Altman, Alan Karr, Urs Gasser, Michael Bar-Sinai, Kobbi Nissim, Jonathan Ullman, Salil Vadhan, and Wojcik, Michael John. 2014. Integrating Approaches to Privacy Across the Research Lifecycle: Long-Term Longitudinal Studies. Social Science Research Network. Cambridge: Harvard University, 07/22. Publisher's VersionAbstract

On September 24-25, 2013, the Privacy Tools for Sharing Research Data project at Harvard University held a workshop titled "Integrating Approaches to Privacy across the Research Data Lifecycle." Over forty leading experts in computer science, statistics, law, policy, and social science research convened to discuss the state of the art in data privacy research. The resulting conversations centered on the emerging tools and approaches from the participants’ various disciplines and how they should be integrated in the context of real-world use cases that involve the management of confidential research data.

This workshop report, the first in a series, provides an overview of the long-term longitudinal study use case. Long-term longitudinal studies collect, at multiple points over a long period of time, highly-specific and often sensitive data describing the health, socioeconomic, or behavioral characteristics of human subjects. The value of such studies lies in part in their ability to link a set of behaviors and changes to each individual, but these factors tend to make the combination of observable characteristics associated with each subject unique and potentially identifiable.

Using the research information lifecycle as a framework, this report discusses the defining features of long-term longitudinal studies and the associated challenges for researchers tasked with collecting and analyzing such data while protecting the privacy of human subjects. It also describes the disclosure risks and common legal and technical approaches currently used to manage confidentiality in longitudinal data. Finally, it identifies urgent problems and areas for future research to advance the integration of various methods for preserving confidentiality in research data.

More

Presentations