BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Can we automatically anonymize text documents? - Pierre Lison\, No
 rwegian Computing Center
DTSTART:20220519T100000Z
DTEND:20220519T110000Z
UID:TALK174584@talks.cam.ac.uk
CONTACT:Marinela Parovic
DESCRIPTION:Text documents often contain personal data in some form. To pr
 otect the privacy of the individuals referred to in those documents\, it i
 s often desirable (and\, in many cases\, mandatory) to edit those document
 s such as to conceal the identity of those individuals. This anonymization
  process remains a difficult task\, at the intersection of NLP\, law and d
 ata privacy. In this talk\, I’ll give an overview of current approaches 
 and outline a number of unsolved problems. Furthermore\, I’ll present th
 e Text Anonymization Benchmark (TAB)\, a new corpus and evaluation framewo
 rk dedicated to this task. TAB contains 1268 court cases from the European
  Court of Human Rights manually enriched with detailed annotations regardi
 ng the personal data expressed in each document.  We hope this new benchma
 rk will inspire NLP researchers to work on this challenging but important 
 problem.
LOCATION:https://cam-ac-uk.zoom.us/j/97599459216?pwd=QTRsOWZCOXRTREVnbTJBd
 XVpOXFvdz09
END:VEVENT
END:VCALENDAR