[gnso-whois-study] FW: Agenda item 6 :: area 8

["Gomes, Chuck" <cgomes@xxxxxxxxxxxx>]

Below are hypotheses proposed by Eric Brunner-Williams for Area 8, Other
GAC Recommendations.  I also entered these into the wiki table.  Note
that it each case he provided two alternative wordings.

Chuck

-----Original Message-----
From: Eric Brunner-Williams [mailto:ebw@xxxxxxxxxxxxxxxxxxxx] 
Sent: Sunday, July 20, 2008 6:19 PM
To: Gomes, Chuck
Subject: Agenda item 6 :: area 8

GAC bullet #1: compile data that provides a documented evidence base
regarding the amount and source of traffic accessing WHOIS servers and
the types and numbers of different groups of users and what those users
are using WHOIS data for.


The hypothesis is that the amount and source of WHOIS requests is
non-uniform, and that the non-uniformity is interpretable, and a likely
interpretation is that data point clusters will corollate with intent,
in a reference data set.

Restated, that some subset of Registrar raw data, minimally containing
just the date and source of WHOIS request data, is disclosed to a
third-party for data analysis, and that use cases for WHOIS data may be
discerned from clusters in a reference WHOIS access data set, and that
the use cases and statistical corollates are disclosed to the GAC.


GAC Recommendation # 5: What is the
percentage of domain name registrants who are natural persons versus
legal persons (or entities)?


The hypothesis is that data exists, or may be captured, which
distinguishes registrants as either "natural persons" or "juristic
persons", and sufficient data exists to infer the percentage of each,
for some unstated subset of all domain registrants.

Restated, that some subset of Registrar raw data, minimally containing
just the distinguishing binary-valued data, is disclosed to a
third-party for data analysis, and the percentage for each value is
disclosed to the GAC.


GAC Recommendation # 6: What is the
percentage of domain name registrations that are registered for and/or
are used for commercial purposes versus those registered for
non-commercial or personal use?  If possible, the data should be broken
down by geographic (e.g. by continent) locations.


The hypothesis is that data exists, or may be captured, which
distinguishes registrations as either "commercially purposed" or
"non-commercially purposed" or "personal use purposed", at some prior
point of time, or the present, or both, and sufficient data exists to
infer the percentage of each, for some unstated subset of all domain
registrations.

Restated, that some subset of Registrar raw data, minimally containing
just the Registrant disclosed tri-valued purpose and a geographic
identifier, or some data from some other provenance, containing the
Registrant's present tri-valued purpose, disclosed by the Registrant to
some third-party, or determined by other means other than Registrant
disclosure, is disclosed to a third-party for data analysis, and the
percentage for each value is disclosed to the GAC, optionally corollated
with the geographic identifier.


GAC Recommendation #3: Are technical
measures available that could effectively curtail misuse of data
published on WHOIS databases while preserving legitimate use and open
access to the databases?


The hypothesis is that an admission control mechanism exists which, when
associated with use cases for WHOIS data, and some reference data set,
can be distinguished from the reference data set without the admission
control mechanism and the use case associations, and on only a proper
subset of the use case associations.

Restated, that a testing methodology exists, and a data set exists, for
which WHOIS admission control mechanisms can be tested, which are no
worse than no admission control, for at least one use case.


Time I will never get back.
Eric