Scaling Private Set Intersection to Billion-Element Sets

• Published in: Financial Cryptography and Data Security pp. 195 - 215
• Main Authors: Kamara, Seny, Mohassel, Payman, Raykova, Mariana, Sadeghian, Saeed
• Format: Book Chapter
• Language: English
• Published: Berlin, Heidelberg Springer Berlin Heidelberg
• Series: Lecture Notes in Computer Science
• Subjects: Adversary Structure; Botnet Detection; Custom Implementation; NoSQL Database; Oblivious Transfer
• ISBN: 9783662454718; 3662454718
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-662-45472-5_13

We examine the feasibility of private set intersection (PSI) over massive datasets. PSI, which allows two parties to find the intersection of their sets without revealing them to each other, has numerous applications including to privacy-preserving data mining, location-based services and genomic computations. Unfortunately, the most efficient constructions only scale to sets containing a few thousand elements—even in the semi-honest model and over a LAN. In this work, we design PSI protocols in the server-aided setting, where the parties have access to a single untrusted server that makes its computational resources available as a service. We show that by exploiting the server-aided model and by carefully optimizing and parallelizing our implementations, PSI is feasible for billion-element sets even while communicating over the Internet. As far as we know, ours is the first attempt to scale PSI to billion-element sets which represents an increase of five orders of magnitude over previous work. Our protocols are secure in several adversarial models including against a semi-honest, covert and malicious server; and address a range of security and privacy concerns including fairness and the leakage of the intersection size. Our protocols also yield efficient server-aided private equality-testing (PET) with stronger security guarantees than prior work.