« search calendars

« DIMACS/Northeast Big Data Hub Workshop on Overcoming Barriers to Data Sharing including Privacy and Fairness

DIMACS/Northeast Big Data Hub Workshop on Overcoming Barriers to Data Sharing including Privacy and Fairness

October 23, 2017 - October 24, 2017

Location:

DIMACS Center

Rutgers University

CoRE Building

96 Frelinghuysen Road

Piscataway, NJ 08854

Click here for map.

Organizer(s):

John Abowd, Cornell University

René Bastón, Columbia University

Tal Rabin, IBM

Adam Smith, Boston University

Salil Vadhan, Harvard University

Rebecca Wright, DIMACS

This workshop will bring together computer scientists, legal scholars, social scientists, and consumers of data to understand the extent to which privacy currently limits the sharing of data. Discussions will include but not be limited to research data and will seek to develop standards and best practices that enable new information flows in domains ranging from healthcare to energy. The workshop will also address issues of fairness of data-driven systems and processes, and their connections to privacy. These topics present central challenges for the flow of data within and between government, commercial, and academic institutions data that are crucial to address science, engineering, and policy challenges.

The workshop will address five major themes:

  • Legal perspectives on privacy and the sharing of research data
  • Private data analysis and de-anonymization
  • Algorithmic fairness, disparate impact, and the ethics of data-based policies
  • Government systems and policies for managing research data (including discoverability)
  • Cryptographic tools for sharing and processing secured data.

Related to each theme, there will be at least one tutorial aimed at entire workshop audience. These tutorials will summarize the state of the artin both research and practicerelated to that theme. There will also be one or more sessions on recent research related to each theme. The presentations in these sessions will explain the main ideas in a language accessible to a broad audience, but will include discussion aimed at specialists.

 

Monday, October 23, 2017

8:45 AM - 9:00 AM

Welcome

9:00 AM - 9:30 AM
9:30 AM - 10:00 AM
10:00 AM - 10:30 AM

Secure Multiparty Computation for Scientific Research

Brett Hemenway, University of Pennsylvania

10:30 AM - 11:00 AM

Break

11:00 AM - 12:00 PM
12:00 PM - 1:30 PM

Lunch

1:30 PM - 2:00 PM
2:00 PM - 2:30 PM

Towards Practical Differential Privacy for SQL Queries

Joseph Near, University of California, Berkeley

2:30 PM - 3:00 PM
3:00 PM - 3:30 PM

Parallel Composition Revisited

Chris Clifton, Purdue University

3:30 PM - 4:00 PM

Break

4:00 PM - 5:00 PM
5:00 PM - 5:30 PM
7:00 PM - 9:00 PM

Dinner at Panico's Restaurant

 

Tuesday, October 24, 2017

9:00 AM - 9:30 AM

Development of Usable, Scalable MPC

Mayank Varia, Boston University

9:30 AM - 10:00 AM
10:00 AM - 10:30 AM

Private Collection of Aggregate Statistics at Scale

Henry Corrigan-Gibbs, Stanford University

10:30 AM - 11:00 AM

Break

11:00 AM - 11:30 AM

Rényi Differential Privacy

Ilya Mironov, Google

11:30 AM - 12:00 PM

Structure and Sensitivity in Differential Privacy: Optimal K-norm Mechanisms

Aleksandra Slavkovic, Pennsylvania State University

12:00 PM - 2:00 PM

Lunch

2:00 PM - 3:00 PM

2020 Decennial Census: Formal Privacy Implementation Update

Stephen Clark, United States Census Bureau

Philip Leclerc, United States Census Bureau

William Sexton, United States Census Bureau

3:00 PM - 3:30 PM
3:30 PM - 4:00 PM

Break

4:00 PM - 4:30 PM

Explorations into Algorithmic Fairness

Rafael Pass, Cornell University

4:30 PM - 5:00 PM

Privacy-Preserving Analytics for Correlated Data

Prateek Mittal, Princeton University

 

Participation is open to all, subject to space limitations. Please register if you'd like to attend.

Talks are mostly by invitation, but we will also consider submissions for potential talks. Please send the organizers an e-mail (to rebecca.wright at dimacs.rutgers.edu) if you'd like to be invited to speak with a description (no more than one page) of your planned topic, before October 6, 2017.

Registration for this event is closed.