Proposal for improvement in evaluation of "similar" strings

["Mustala, Tero (NSN - FI/Espoo)" <tero.mustala@xxxxxxx>]

Dear Sirs,

After several discussions in ICANN meetings, the latest in San Francisco
in March, I thought it might be good to bring one small but still
important issue to your attention. It is regarding String Reviews and in
particular  Similarity to Other Applied-for gTLD Strings, as described
in section 2.2.1 of the draft Guidebook. The text in the Guidebook
describes the Review Methology and use of a special algorithm to verify
the similarity. The generic criterion  in evaluation is "user
confusion", and the decision is for the review panel to make.

Let's make a test with two hypothetical applicants and strings, and
assume, that two well known companies submit their applications for the
strings ".abc" and ".bbc" respectively. Result from the similarity
assessment tool is 92%, which I would assume shows high similarity i.e.
high probability to user confusion. The end result of the review might
then be that these applications would be set to a contention set leading
to auction, if no other obstacles come on their way.

What would be the result of the auction? Nobody knows, but we can with
good reasons speculate, that either of the applicants would win and pay
a considerable fee to ICANN, which in turn would need to tell the world
why it has blocked a new gTLD from an established applicant even though
these both strings are used in the ".com domain" without any problems.
Does this result support the general goals of the New gTLD program? I
doubt.

What would be the solution to avoid this kind of problems. All that is
needed is a small addition to the chapter " Similarity to Other
Applied-for gTLD Strings (String Contention Sets)" on page 2-5 of the
Guidebook. That would tell that as ICANN will notify the applicants who
are part of a contention set, it will give the applicants a fixed period
of time to submit  written statements that they do not object the other
applications in this set. In that case the contention set would be
reversed, and the applications could proceed.

This kind of small updates are important, as they may offer solutions to
many potential applicants, who now see big hurdles in the area of string
similarity, and possibly decide not to apply because of high risk of
failure.

Best regards,

Tero Mustala


Tero Mustala
Principal Consultant,
CTO/Industry Environment
Nokia Siemens Networks
tero.mustala@xxxxxxx