BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Towards Statistical Queries over Distributed Private User Data - P
 aul Francis (MPI-SWS)
DTSTART:20120402T150000Z
DTEND:20120402T160000Z
UID:TALK37145@talks.cam.ac.uk
CONTACT:Eiko Yoneki
DESCRIPTION:Today the method du jour for statistical analysis of user beha
 vior is to gather lots of user data\, anonymize it (more-or-less)\, and th
 en analyze that data.  The need for doing statistical analysis drives many
  companies to gather large amounts of user data\, often without the users'
  \nawareness.  My research group at MPI-SWS has been exploring approaches 
 for doing statistical analysis without gathering user data.  Rather\, user
  data is kept on user devices\, and queries are pushed to these \ndevices.
    The resulting answers are anonymized and fuzzed such that 1) \nno singl
 e party can associate data with individual users\, and 2) the aggregate an
 swers are differentially private.  In this talk\, I will present a general
  approach that we will present in NSDI this year.  I will outline the shor
 tcomings of this approach\, and follow with some enhancements that scale b
 etter in specific applications domains\, namely web analytics and behavior
 al advertising.\n\nBio: Paul Francis is a tenured faculty at the Max Planc
 k Institute for Software\nSystems in Germany. Paul has held research posit
 ions at Cornell\nUniversity\, ACIRI\, NTT Software Labs\, Bellcore\,and MI
 TRE\, and was Chief\nScientist at two Silicon Valley startups. Paul’s re
 search centers around\nrouting and addressing problems in the Internet and
  P2P networks. Paul’s\ninnovations include NAT\, shared-tree multicast\,
  the first P2P multicast\nsystem\, the first DHT (as part of landmark rout
 ing)\, and Virtual\nAggregation. Recently Paul has become interested in de
 signing advertising\nsystems that protect user privacy while allowing for 
 effective targeting.
LOCATION:FW26\, Computer Laboratory\, William Gates Builiding
END:VEVENT
END:VCALENDAR