BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Assessing Re-identification Risk in Sample Microdata - Natalie Shl
 omo (University of Manchester)
DTSTART:20161208T163000Z
DTEND:20161208T170000Z
UID:TALK69398@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:Co-author: Chris Skinner &nbsp\;  &nbsp\;  &nbsp\;<br><br>Abst
 ract:  &nbsp\;  &nbsp\;Disclosure risk occurs when there is a high probabi
 lity that an   &nbsp\;&nbsp\;intruder can identify an individual in releas
 ed sample microdata and&nbsp\;confidential information may be revealed. A 
 probabilistic modelling   &nbsp\;framework based on the Poisson log-linear
  model is used for   &nbsp\;quantifying disclosure risk in terms of popula
 tion uniqueness when   &nbsp\;population counts are unknown. This method d
 oes not account for   &nbsp\;measurement error arising either naturally fr
 om survey processes or   &nbsp\;purposely introduced as a perturbative dis
 closure limitation technique.  The probabilistic modelling framework for a
 ssessing disclosure risk is   &nbsp\;expanded to take into account the mis
 classification/ perturbation and   &nbsp\;demonstrated on sample microdata
  which has undergone &nbsp\;&nbsp\;perturbation   &nbsp\;procedures. Final
 ly\, we adapt the probabilistic modelling framework to   &nbsp\;&nbsp\;ass
 ess the disclosure risk of samples from sub-populations   &nbsp\;and show 
 some initial results.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR