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

Tue May 23 14:19:49 UTC 2017

Given the tension between those two paragraphs, I'm assuming that
Brian meant "the false positive rate doesn't have to increase a lot."
To evaluate this claim, we'd need a proposal for some limits rather
than hypothesizing their existence--has anyone prepared such a
proposal? The one actual proposal we have doesn't have limits beyond
"string included."  I also think the "common terms" proposal has
significant weaknesses, such as triggering match notices for
"hoteljobs" and the like, given what we know about what's in the TMCH.

Brian's statement about evaluating matches also makes more sense from
the perspective of a trademark owner receiving a report from a watch
service, the current process, rather than a recipient of a match
notice who may be trying to register one domain name and, especially
if they're trying to register the name of their own hotel, may not
have the relevant background.
Rebecca Tushnet
Georgetown Law
703 593 6759

On Tue, May 23, 2017 at 9:34 AM, Winterfeldt, Brian J.
<BWinterfeldt at mayerbrown.com> wrote:
> Dear all:
>
>
>
> Here are a couple of important points to consider regarding false positives,
> which are largely a red herring in my opinion.
>
>
>
> First, any brand owner or representative that uses commercial watch services
> will tell you that false positives will always be present, no matter how
> good the algorithm used to flag matters of potential concern.  Predicates
> for enforcement measures, including likelihood of confusion for infringement
> and bad faith for cybersquatting, are analyses that are necessarily made by
> trained professionals.
>
>
>
> Second, as was pointed out on our last call, expanded TMCH matches need not
> wildcard, or be compared against, the entire universe of possible terms
> (though that approach would certainly make sense for fanciful marks).  For
> example, relevant terms could be taken from the goods and services listed in
> the TMCH record, or from a predetermined list of commonly abused terms like
> CAREERS, JOBS, and HOME etc.  Thus, the system could be designed in a way
> such that there would be no false positives.
>
>
>
> Thank you,
>
>
>
> Brian
>
>
>
> Brian J. Winterfeldt
>
> Co-Head of Global Brand Management and Internet Practice
>
> Mayer Brown LLP
>
>
>
> -----Original Message-----
> From: gnso-rpm-wg-bounces at icann.org [mailto:gnso-rpm-wg-bounces at icann.org]
> On Behalf Of icannlists
> Sent: Thursday, May 18, 2017 2:46 PM
> To: Rebecca Tushnet
> Cc: gnso-rpm-wg at icann.org
> Subject: Re: [gnso-rpm-wg] Data on non exact matches
>
>
>
> Rebecca,
>
>
>
> Apologies for copying the wrong list before. I've corrected it here.
>
>
>
> I agree wholeheartedly that we need to minimize false positives. I'm looking
> forward to rolling up our collective sleeves to find the right balance!
>
>
>
> Best,
>
> Paul
>
>
>
>
>
> Sent from my iPhone
>
>
>
>> On May 17, 2017, at 7:10 PM, Rebecca Tushnet <rlt26 at georgetown.edu> wrote:
>
>>
>
>> My point was instead that even if the best data weren't collected there
>> are proxies. And I wasn't talking about chilling effects, though those are
>> relevant, but about false positives, which are detrimental to the system as
>> a whole, including to legitimately sent notices.
>
>>
>
>> Sent from my phone. Apologies for terseness/typos.
>
>>
>
>>> On May 17, 2017, at 7:13 PM, icannlists <icannlists at winston.com> wrote:
>
>>>
>
>>> Understood Rebecca, but we also can't just shrug off demonstrable harm to
>>> brands and their consumers just because we are data poor on the drive to
>>> identify a chilled would-be registrant.
>
>>>
>
>>> Even so, hopefully, Staff will respond soon so we can know for sure
>>> whether or not the data you are seeking was collected.
>
>>>
>
>>>
>
>>> Best,
>
>>> Paul
>
>>>
>
>>>
>
>>>
>
>>> -----Original Message-----
>
>>> From: Rebecca Tushnet [mailto:Rebecca.Tushnet at law.georgetown.edu]
>
>>> Sent: Wednesday, May 17, 2017 5:13 PM
>
>>> To: icannlists <icannlists at winston.com>
>
>>> Subject: Re: [gnso-rpm-wg] Data on non exact matches
>
>>>
>
>>> As George has been saying, there are ways to estimate false positives
>>> using various datasets.  We shouldn't go ahead considering only claims about
>>> benefits and just shrugging about costs and promising to look into them
>>> later; that seems contrary to the reasons exact match was adopted as the
>>> consensus at the outset.
>
>>> Rebecca Tushnet
>
>>> Georgetown Law
>
>>> 703 593 6759
>
>>>
>
>>>
>
>>>> On Wed, May 17, 2017 at 6:09 PM, icannlists <icannlists at winston.com>
>>>> wrote:
>
>>>> Thanks Rebecca,
>
>>>>
>
>>>> The point I was trying to get at is that even if we are dataless on the
>>>> false positives from Round 1 (harm to potentially chilled registrants) we
>>>> aren't dataless on the harm to trademark owners and the consumers they
>>>> protect.  Say we pick a mark and add to it the product terms contained in
>>>> its statement of services, we can then check that against the URS and UDRP
>>>> records for new gTLDs to see if they would have triggered a claims notice.
>>>> So that part, at least, is knowable (although quite the job for anyone
>>>> trying to dig into the data).
>
>>>>
>
>>>> I'm looking forward to Staff's reply on whether or not there is a list
>>>> of domain names abandoned after claims notice.  I'm pretty sure the answer
>>>> is "no" but let's be safe.  If no, then we have to ask should we continue to
>>>> allow the harms evidence by UDRP & URS decisions just because we forgot to
>>>> collect data on abandonments following claims, or so we rely on the UDRP &
>>>> URS analysis to say we need to enhance claims (and then collect the data in
>>>> the second round to see if the enhanced claims need further tweaking either
>>>> up or down).
>
>>>>
>
>>>> Hopefully, we will hear from Staff soon.
>
>>>>
>
>>>> Best,
>
>>>> Paul
>
>>>>
>
>>>>
>
>>>>
>
>>>> -----Original Message-----
>
>>>> From: Rebecca Tushnet [mailto:Rebecca.Tushnet at law.georgetown.edu]
>
>>>> Sent: Wednesday, May 17, 2017 5:01 PM
>
>>>> To: icannlists <icannlists at winston.com>
>
>>>> Subject: Re: [gnso-rpm-wg] Data on non exact matches
>
>>>>
>
>>>> Paul, I don't think that the idea that UDRP/URS decisions are relevant
>>>> evidence is true across the board; even after excluding those with
>>>> common-law rights only, we'd still have to look at which ones would have
>>>> gotten a notice if the match process had been in place.
>
>>>>
>
>>>> Also, consumers have interests in the websites of legitimate
>>>> registrants; consumer interests do not line up exactly with trademark
>>>> owners' interests.
>
>>>>
>
>>>> Separately, the percentage of false positives is important to trademark
>>>> owners too; if 80% of attempts trigger a match notice, that is likely to
>>>> diminish the deterrent effect--the way we click through a number of other
>>>> warnings that don't seem particularly targeted to our activity.
>
>>>> Rebecca Tushnet
>
>>>> Georgetown Law
>
>>>> 703 593 6759
>
>>>>
>
>>>>
>
>>>>> On Wed, May 17, 2017 at 5:48 PM, icannlists <icannlists at winston.com>
>>>>> wrote:
>
>>>>> Thanks Rebecca.
>
>>>>>
>
>>>>> Staff, does the information that Rebecca is seeking in existence?  I
>>>>> thought that there was no record of the specific domain names which were
>>>>> non-registered after receiving a notice of claim.
>
>>>>>
>
>>>>> Rebecca, I think your question has more to do with harm to potentially
>>>>> chilled registrants rather than harm to trademark owners/consumers, correct?
>>>>> Harm to trademark owners/consumers can be evidenced by UDRP/URS decisions
>>>>> for domain names that would fall into the next expanded category.  Thanks in
>>>>> advance for your thoughts on that.
>
>>>>>
>
>>>>> Best,
>
>>>>> Paul
>
>>>>>
>
>>>>>
>
>>>>>
>
>>>>> -----Original Message-----
>
>>>>> From: gnso-rpm-wg-bounces at icann.org
>
>>>>> [mailto:gnso-rpm-wg-bounces at icann.org] On Behalf Of Rebecca Tushnet
>
>>>>> Sent: Wednesday, May 17, 2017 1:36 PM
>
>>>>> To: gnso-rpm-wg at icann.org
>
>>>>> Subject: [gnso-rpm-wg] Data on non exact matches
>
>>>>>
>
>>>>> Reposting per J. Scott's request on the call:
>
>>>>>
>
>>>>> In order to fully evaluate the proposal to extend the match criteria,
>>>>> we really need information on the harm it is supposed to address, including
>>>>> (1) the number of registration attempts that included but were not limited
>>>>> to the text of an entry in the TMCH and (2) the number of such registrations
>>>>> that led to a trademark owner challenge after the fact (therefore excluding
>>>>> URS actions against, for example, typosquatting).
>
>>>>> To the extent that the proposal becomes more limited to specific goods
>>>>> 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.
>
>>>>>
>
>>>>>
>
>>>>>
>
>>>>>
>
>>>>> Rebecca Tushnet
>
>>>>> Georgetown Law
>
>>>>> 703 593 6759
>
>>>>> _______________________________________________
>
>>>>> gnso-rpm-wg mailing list
>
>>>>> gnso-rpm-wg at icann.org
>
>>>>>
>>>>> https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmm.icann.org%2Fmailman%2Flistinfo%2Fgnso-rpm-wg&data=01%7C01%7CPMarano%40mayerbrown.com%7C297e181615e543123b5508d49e1e3280%7C09131022b7854e6d8d42916975e51262%7C0&sdata=TurMe1MF7qc5EFxuIUrgsC0OMKUwVly8w2zMB%2FAERm4%3D&reserved=0
>
>>>>>
>
>>>>> ________________________________
>
>>>>> The contents of this message may be privileged and confidential. If
>>>>> this message has been received in error, please delete it without reading
>>>>> it. Your receipt of this message is not intended to waive any applicable
>>>>> privilege. Please do not disseminate this message without the permission of
>>>>> the author. Any tax advice contained in this email was not intended to be
>>>>> used, and cannot be used, by you (or any other taxpayer) to avoid penalties
>>>>> under applicable tax laws and regulations.
>
> _______________________________________________
>
> gnso-rpm-wg mailing list
>
> gnso-rpm-wg at icann.org
>
> https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmm.icann.org%2Fmailman%2Flistinfo%2Fgnso-rpm-wg&data=01%7C01%7CPMarano%40mayerbrown.com%7C297e181615e543123b5508d49e1e3280%7C09131022b7854e6d8d42916975e51262%7C0&sdata=TurMe1MF7qc5EFxuIUrgsC0OMKUwVly8w2zMB%2FAERm4%3D&reserved=0
>
> __________________________________________________________________________
>
>
> This email and any files transmitted with it are intended solely for the use
> of the individual or entity to whom they are addressed. If you have received
> this email in error please notify the system manager. If you are not the
> named addressee you should not disseminate, distribute or copy this e-mail.
>
>
>
> _______________________________________________
> gnso-rpm-wg mailing list
> gnso-rpm-wg at icann.org
> https://mm.icann.org/mailman/listinfo/gnso-rpm-wg