[Corpora-List] Anonymisation (via BNC)

Miles Osborne miles at inf.ed.ac.uk
Mon Aug 18 10:44:21 UTC 2014


I would be surprised if the BNC could be used for building an automatic
anon model.  Apart from anything else, you would need the non-anonymised
version so you know what was and wasn't anonymised.

But in general anonymisation is a fairly weak form of privacy protection
since (a) you need to know in advance what you are protecting and (b) it
makes no guarantees such as being resistant to linkage attacks.

http://arxiv.org/abs/cs/0610105

(netflix anonymisation broken)

Miles

-- 
The University of Edinburgh is a charitable body, registered in Scotland,
with registration number SC005336.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20140818/96e58bb1/attachment.htm>
-------------- next part --------------
_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora


More information about the Corpora mailing list