[gnso-rpm-wg] Data on non exact matches

[George Kirikos]

Hi folks,

On Wed, May 17, 2017 at 2:35 PM, Rebecca Tushnet
<Rebecca.Tushnet at law.georgetown.edu> wrote:
> and services, the evidence of need should be likewise limited.  Then,
> as George K. eloquently explained, we'd also want to see how many
> attempts/registrations would have been flagged in an expanded match
> system, to get a sense of the potential for false positives.

(1) Just to expand on this while it's still fresh in my mind, one can
run a simulation using a proposed expanded matching rule, to test what
the increased number of claims notices would be. If, for example, the
frequency of claims notices increased from 30% to 31%, that would be a
more acceptable situation compared to an increase from 30% to 80% or
90%.

To run those simulations, one would need an appropriate data set on
historical *attempted* registrations. Or, one would have to start
collecting that data, if it's not already being done. One could, in
the alternative, run it against the .com zone file, or another similar
data set, but that'd be imperfect (since the .com zone file never had
claims notices to begin with, doesn't match *attempted* registrations,
etc.).

(2) For those advocating for "expanded notices", a public TMCH
database would allow good faith registrants to do their own checks for
plurals, or other types of "expanded matches" on their own.

(3) As I noted during the call, for longer domain name registrations,
there could be multiple TM claims notices for a single domain name.

e.g. MAGMA.TLD could trigger:

(i) MAGMA -- http://tsdr.uspto.gov/#caseNumber=86888435&caseType=SERIAL_NO&searchType=statusSearch
(ii) GM -- General Motors, obviously
(iii) MAG -- http://tsdr.uspto.gov/#caseNumber=87065941&caseType=SERIAL_NO&searchType=statusSearch
(iv) AG -- http://tsdr.uspto.gov/#caseNumber=85293626&caseType=SERIAL_NO&searchType=statusSearch
(v) AGM -- http://tsdr.uspto.gov/#caseNumber=77181108&caseType=SERIAL_NO&searchType=statusSearch
(vi) GMA -- http://tsdr.uspto.gov/#caseNumber=75268825&caseType=SERIAL_NO&searchType=statusSearch
(and maybe others!)

and that was just for an innocuous 5-letter domain. Imagine the
matches for a 10 or 15 character domain? Registrants might need to
check 15, 20, or more TM claims notices, before being allowed to
register a domain. The TM+50 trade-off seemed to recognize this
situation.

(4) Another solution, although it has issues amongst some TM holders,
would be to limit expanded matches only to those marks that are most
damaged by cybersquatting, i.e. auction off a limited number (say 50
or 100) of slots that get the expanded notice privilege. Financial
firms like Paypal, or other firms that are often the targets of
cybersquatting, e.g. Microsoft, Google, Yahoo would likely be the
winners of those auctions. The auctions could be held every 6 months
or a year, so the market would determine the "value" of those slots
(e.g. a firm that sees it doesn't get much benefit wouldn't bid as
much the next time, and some other firm would win the slot, etc.).

This solution recognizes the reality that not all marks are equal in
strength or in economic value, or are equally at risk for
cybersquatting. It avoids defining what is a so-called "famous" mark,
by allowing the market mechanism (auctions) to allocate the slots.
With strict limits as to the number of slots, it also helps reduce the
"false positives" above (i.e. the simulation in the first paragraph of
point 1 above wouldn't change as much when it's only 50 or 100 marks,
compared to the entire TMCH database of 40,000+ marks).

Proceeds from such a routine auction might also be allocated to causes
related to reducing reverse domain name hijacking (to directly balance
against the expansion of TM claims notices).

Sincerely,

George Kirikos
416-588-0269
http://www.leap.com/