[Colloquium] Xu/MS Presentation/Jun 12, 2017
Margaret Jaffey via Colloquium
colloquium at mailman.cs.uchicago.edu
Thu May 25 10:22:39 CDT 2017
This is an announcement of Min Xu's MS Presentation.
------------------------------------------------------------------------------
Date: Monday, June 12, 2017
Time: 10:30 AM
Place: Ryerson 277
M.S. Candidate: Min Xu
M.S. Paper Title: HERMETIC: PRIVACY-PRESERVING DISTRIBUTED ANALYTICS
WITHOUT (MOST) SIDE CHANNELS
Abstract:
Distributed analytics systems, such as Spark, enable users to
efficiently perform computations over large distributed data sets.
Recently, a number of systems have been proposed that can additionally
protect the privacy of the data, by keeping it encrypted even in
memory, and by performing the computations using trusted hardware
features, such as Intels SGX. This approach is attractive because it
makes it much safer to outsource computations, e.g., to an untrusted
cloud platform. However, existing solutions remain vulnerable to a
variety of side channels, such as timing, message sizes, and cache
contents, which considerably weakens the privacy guarantees that these
systems can effectively provide. In this paper, we present a
principled approach to closing or mitigating the most critical side
channels. We introduce a new primitive that can perform simple
computations on a locked-down core, without externally visible side
effects, as well as several enhanced obliv- ious algorithms that can
use this primitive to answer more complex queries over large data
sets. To preserve efficiency and to enable optimizations, we allow our
algorithms to carefully disclose a small amount of information via
differentially private queries. We present the design of a system
called Hermetic that uses our techniques to answer SQL-style queries;
our experimental evaluation of a Hermetic prototype shows that it is
competitive with previous privacy-preserving systems, even though it
provides stronger privacy guarantees.
Min's advisor is Prof. Ariel Feldman
Login to the Computer Science Department website for details:
https://www.cs.uchicago.edu/phd/ms_announcements#xum
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Margaret P. Jaffey margaret at cs.uchicago.edu
Department of Computer Science
Student Support Rep (Ry 156) (773) 702-6011
The University of Chicago http://www.cs.uchicago.edu
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
More information about the Colloquium
mailing list