W3C home > Mailing lists > Public > public-privacy@w3.org > October to December 2010

Re: do not track list?

From: Karl Dubost <karld@opera.com>
Date: Wed, 17 Nov 2010 08:46:49 -0500
Message-Id: <7CEA48B3-7AB6-4D74-A045-D931DEA827C8@opera.com>
Cc: "Thomas Roessler" <tlr@w3.org>, "Rigo Wenning" <rigo@w3.org>, <public-privacy@w3.org>
To: "Chappelle, Kasey, VF-Group" <Kasey.Chappelle@vodafone.com>

Le 17 nov. 2010 à 07:49, Chappelle, Kasey, VF-Group a écrit :
apply to aggregate tracking - the kind of
> statistical analysis that does not capture individual profiles, but does
> do some kind of tracking at the very lowest level to create those
> statistics (unique visitors, for example). Some regulators already
> consider this too to be privacy-invasive

Because the issue is not about tracking per say. This is the wrong end of the stick. The issue is more on the context of tracking. 

If one tracks my position when 

Case 1:
	* I'm in a very crowded city 
	* during rush hours 
	* with a precision of 200m.

Case 2:
	* I'm in the country side (nobody around but me)
	* during week days
	* with a precision of 200m

In case 1, I could say yes you can track me, there is not that much issues in the aggregate because my position will be drowned in hundred of other overlapping positions. A cloud of probability. The opacity is thicker.

In case 2, with exactly the same settings, the opacity is thin, very thin, to inexistent.

Same kind of variations of data will happen for example if you consider. You can track me once, or you can track me everyday. The density of information will make possible to identify the person in one place. Ex: anonymous position of people vers id position of people. With the id, you can figure out the habits, the home, the work place, etc.

-- 
Karl Dubost - http://dev.opera.com/
Developer Relations & Tools, Opera Software
Received on Wednesday, 17 November 2010 13:47:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 17 November 2010 13:47:29 GMT