UMass Amherst College of Information and Computer Sciences (CICS) professor Gerome Miklau, associate professor Andrew McGregor and alumni Chao Li and Michael Hay are scheduled to receive the Alberto O. Mendelzon Test-of-Time Award at the 2020 Association for Computing Machinery Symposium on the Principles of Database Systems (ACM PODS) for their 2010 paper, “Optimizing Linear Counting Queries under Differential Privacy.”
The paper, co-authored with Vibhor Rastogi, then a doctoral student at the University of Washington, introduced a novel method for answering multiple related queries accurately, in the emerging model of differential privacy, in which queries are answered “noisily” to protect personal information. Their new algorithm, which they called the matrix mechanism, significantly decreased the amount of noise added to results by automatically creating a different set of strategically chosen queries, whose answers are then systematically used to derive answers to the original queries. Their approach provided a theoretical design for an automated procedure—which has since been put into practice—to automatically improve overall accuracy and allow more important queries to be prioritized for greater accuracy.
The 2020 Award Committee cited the paper’s significant impact, “This matrix mechanism has been fundamental to further developments in the field. Most notably is the adoption of the approach as the data-privacy mechanism in the United States 2020 Census. The paper received about 300 citations, both in database theory and data systems papers, underscoring its additional impact. Furthermore, it received numerous nominations from the community.”
ACM grants the Alberto O. Mendelzon Test-of-Time Award every year to a single paper or small number of papers published in the PODS proceedings at least ten years prior that have had “the most impact in terms of research, methodology, or transfer to practice over the intervening decade.” The researchers this year were recognized for their contributions in the field of differential privacy, especially for their advancements in data-disclosure-avoidance applications.
Gerome Miklau joined the faculty of CICS in 2005. His research focuses on private and secure data management. He designs algorithms to accurately learn from data without disclosing sensitive facts about individuals, primarily in the model of differential privacy. He also designs novel techniques for controlling access to data, limiting retention of data, and resisting forensic analysis.
Andrew McGregor joined the faculty of CICS in 2009. His research includes algorithms for processing massive data sets and data streams, computing with noisy or incomplete data, clustering, communication complexity, and coding and information theory.
Chao Li is a software engineer at Google, and received his PhD in computer science from the University of Massachusetts Amherst.
Michael Hay is an associate professor of computer science at Colgate University and co-founder, with Miklau, of Tumult Labs. He received his PhD from the University of Massachusetts Amherst in 2010.