As we continue our work to improve privacy at Facebook, one of the key elements is learning from outside experts. We’re doing this across many different disciplines. Today, we want to share the results of some initial grants as we start to build academic research relationships that will let us start better funding the types of privacy technology that everyone can benefit from in the future.

This past November, Facebook launched a new research award opportunity in privacy-preserving technology at the 2019 ACM Conference on Computer and Communications Security. This RFP was the first of a series of privacy-related research award opportunities where we look to engage with academia worldwide. Soon after that, in January, we launched a second research award opportunity, which was about the role of applied cryptography in a privacy-focused advertising ecosystem.

After a thorough internal review process, the winners and finalists of these RFPs have been selected. We received a total of 164 research proposals and have selected 14 winners: six for privacy-preserving technology and eight for the role of applied cryptography. Facebook research teams in infrastructure, UX, marketing science, advertising, blockchain, and machine learning took part in the review process.

“Selecting winners from this impressive set of proposals was not easy,” says Sharon Ayalde, Research Program Manager, Academic Relations. “We look forward to seeing the outcome of the research described in these outstanding proposals and to collaborating further with these experts.”

In our continued goal of bridging the gap between industry and academia, the winners of both RFPs are also invited to a Privacy Faculty Summit in Menlo Park (timing TBD), where they will be able to connect with Facebook researchers, discuss their research, and network with one another. As a continuation of our investment in privacy research, we will launch another RFP in August 2020 — this time with a focus on UX privacy challenges.

“As Facebook and the industry overall move into this modern era of privacy, we think there are many interesting, unsolved technical privacy challenges that require collaboration across industry and academia to find rigorous solutions,” says Scott Renfro, Facebook Software Engineer.

We appreciate the time everyone took to submit a proposal. To receive email notifications about upcoming research awards, subscribe to our RFP newsletter.

Research award winners

Principal investigators are listed first unless otherwise noted.

Privacy-preserving technologies

Achieving deletion guarantees in LSM data stores
Manos Athanassoulis (Boston University)

Differentially private SQL: Privacy verification and accuracy optimization
Danfeng Zhang, Daniel Kifer (Pennsylvania State University)

In-situ privacy controls of profiling and ad-targeting
Yang Wang (University of Illinois)

Private user authentication: Anonymous credentials on Facebook’s Libra blockchain
Foteini Baldimtsi (George Mason University), Anna Lysyanskaya

Provably-secure kNN search
Alexandra Boldyreva, Tianxin Tang (Georgia Institute of Technology)

Transparency.js, Transparency for active content
Michael Backes, Ilkan Esiyok, Robert Kuennemann (CISPA Helmholtz Center for Information Security)

Role of applied cryptography in a privacy-focused ecosystem

A general pragmatic approach to privacy-preserving advertising ecosystem
Vladimir Kolesnikov (Georgia Institute of Technology)

Cryptographic tools for privacy-friendly advertising
Henry Corrigan-Gibbs (Massachusetts Institute of Technology)

Differentially private deep learning with applications to online advertising
Weijie Su (University of Pennsylvania), Xi Chen (New York University)

Flexible and scalable secure measurement aggregation
Taeho Jung (University of Notre Dame)

Know your anonymous customer
Anna Lysyanskaya (Brown University)

Next-generation private record linkage
Mike Rosulek (Oregon State University)

Privacy-preserving multi-party sketching for advertisement measurement
Arkady Yerukhimovich (George Washington University)

Private integration of retailer’s data feeds in an advertising ecosystem
Olya Ohrimenko (University of Melbourne)

Research award finalists

Privacy-preserving technologies

Accuracy-aware differential privacy for multi-user systems
Xie He (University of Waterloo)

Efficient privacy-preserving autoML system for content-based recommendation
Bo Li (University of Illinois), Ahmad Beirami (Facebook), Pengtao Xie (University of California San Diego), Yanzhi Dou (Facebook)

Encrypted search from theory to browsers
David Cash (University of Chicago)

End-to-end encryption for collaboration software
Martin Kleppmann, Alastair R. Beresford (University of Cambridge)

Good explanation for algorithmic transparency in targeted ads
Dokyun “DK” Lee, David Danks, Joy Tong Lu, Taewan Kim (Carnegie Mellon University)

Locally private analysis of graphs
Sofya Raskhodnikova (Boston University)

Preserving privacy: Quantifying the gap in perceived and actual privacy risks
Phillip Morgan, Emily Collins, Dylan Jones, Pete Burnap, Tasos Spiliotopoulos (Cardiff University)

Privacy-aware targeted advertising
Dokyun “DK” Lee (Carnegie Mellon University)

Privacy methods and tools for Arab world based on cultural and Islamic values
Karen Elizabeth Fisher (University of Washington), Eaid Yafi (University of Kuala Lumpur), Yacine Ghamri-Doudane (University of La Rochelle)

Privacy-preserving data analysis for social processes
Zhiwei Steven Wu (University of Minnesota), Aaron Schein (Columbia University)

Split learning: Privacy-preserving efficient distributed machine learning
Ramesh Raskar (Massachusetts Institute of Technology)

Streaming data analytics under local differential privacy
Ninghui Li (Purdue University)

Subversive machine learning: Countering surveillance with usable obfuscation
Sauvik Das (Georgia Institute of Technology)

Role of applied cryptography in a privacy-focused ecosystem

Decentralized federated leaning for safer personalized advertising
Yan Huang (Kennesaw State University), Zhipeng Cai (Georgia State University)

Privacy-preserving tailored advertising
Anderson Nascimento (University of Washington)

Private and robust record linkage using record-level Bloom filters
Marat Kantarcioglu (University of Texas at Dallas)

TEFLA: Tailor-made encryption meets federated learning applications
Adam O’Neill (University of Massachusetts Amherst)

Zero-knowledge proofs for cross-business purchase measurement
Yupeng Zhang (Texas A&M University), Charalampos Papamanthou (University of Maryland, College Park)