Re: Demographic Data Collection (was Re: Automatic Entry and Forms)

hallam@zorch.w3.org
Wed, 28 Feb 96 12:21:56 -0500


From: hallam@zorch.w3.org
Message-Id: <9602281721.AA18477@zorch.w3.org>
To: Brian Behlendorf <brian@organic.com>
Cc: hallam@zorch.w3.org, www-html@w3.org
Subject: Re: Demographic Data Collection (was Re: Automatic Entry and Forms)  
In-Reply-To: Your message of "Wed, 28 Feb 96 02:10:20 PST."
             <Pine.SGI.3.91.960228001333.11520p-100000@fully.organic.com> 
Date: Wed, 28 Feb 96 12:21:56 -0500


Brian makes the usefull point that giving away data is a type of
payment. I'm not sure I entirely agree with his assertion that
this means we should consider demographic data exchange as a payment 
system rather thab as an application level system however.

I see the transition between the payments and the application layer
being made when we are dealing with authenticated rather than casual
demographic profiles. I made some hints about this work in my proxy caching 
paper where I introduce the idea of a data escrow agent which a server would 
send a log file to and recieve back an analysis using aggregated demographic 
classes (middle aged burnouts, young and useless, old farts, etc...)


I was wondering if we should approach the privacy question politically
rather than simply technologically. Whatever proposal is made there will
be people who find some way of abusing it. There will also be very good
reasons for a customer to give sensitive data to a server. Maybe what we
need is a code of practice with a trademarked "seal of approval" stating
that the site abides by certain confidentiality commitments.


I very much like Brian's suggestion of using SGML entities as short forms
for local data. &downloaded-date; and similar transport related attributes come 
to mind as well :-) 


		Phill