W3C home > Mailing lists > Public > public-tracking@w3.org > February 2013

Swire notes to discuss first break-out session

From: Peter Swire <peter@peterswire.net>
Date: Mon, 11 Feb 2013 12:54:04 -0800
To: "public-tracking@w3.org (public-tracking@w3.org)" <public-tracking@w3.org>
Message-ID: <CD3EC46D.6E13B%peter@peterswire.net>
Read out from 1st Breakout

difficult history of DNT: 18 months of inter-personal history
            define important use cases with time limits (hard/soft time limits)


1st party/3d party
            longer uses 1st party; targeting and often shorter time frame for that use case
            across multiple sites
            user can delete from 1st party; portability/manage

definitions

“a reasonable portion over a significant time”  (David Singer)


 (Nick Doty)
1. specificity of data; full URL (also have query parameters such as search terms) vs. domain/path (nytimes.com/article name) vs. just domain; extracted category

2. how associated to the user: not browsing history if not linked back to a user; user ID, deID layer (longitudinal but not connected back) or aggregate

3.  time: delete data by an amount of time; no agreement on those levels; session


context of three kinds of data sets (Heather West):
            1. raw data – event level transaction data; clickstream and also lots more; background stuff
            2. managed data – may have been enriched/appended, been processed in some way but is linkable, cookie to a device or
            3. de-ID; hash the identifier and throw away the salt; is same as rotating a hash

            closer to raw data and stricter controls organizationally


Data minimization:
            should be able to demonstrate the need; and not show anything else; maximum retentionperiod disclosed

            time limit: reasonable vs. categories for uses

antifraud:
            don’t get on Safari now
            role of IP address rather than full URI
            cost/impression; cost/click; cost/action


for targeting – if go to bucket, what if convert immediately to car enthusiast
            pseudonymous required there to make sense


things for further fact development:
            financial audit
            cybersecurity
            anti-fraud


rotate a hash; deID for tomorrow; quantum computers poses a problem



Professor Peter P. Swire
C. William O'Neill Professor of Law
    Ohio State University
240.994.4142
www.peterswire.net
Received on Monday, 11 February 2013 20:54:51 UTC

This archive was generated by hypermail 2.3.1 : Friday, 21 June 2013 10:11:42 UTC