[NCSG-PC] Internationalized Domain Names and string similarity reviews

[Tomslin Samme-Nlar]

Dear NCSG,

The Internationalized Domain Names (IDN) EPDP team is currently discussing
charter questions on how string similarity should be performed on variant
labels of future TLD labels.
3 options described as *levels* were proposed as you will see on the
attached slide deck and the team is to choose one of them. I'll try to
summarize the description of the 3 levels below and indicate my preferred
level.

*Notes for those not familiar with the topic:*




*In IDNs, a TLD has a primary label (P) and variants (v), which are labels
in different languages and/or scripts of the primary label.Consider P1 as
the applied-for primary gTLD 1 and P1 could have variant labels, say three
for the sake of this discussion:○ P1v1: allocatable and requested for
activation○ P1v2: allocatable but not requested○ P1v3: blocked*

Now, the question is, during an application round, for string similarity
reviews, should P with its variants P1v1, P1v2 & P1v3 be compared with

   - *Level 1*:* Existing & applied for Primary labels + ONLY Requested
   Allocatable Variants*
      - *Pros*: (a) Limited pool of labels for comparison; (b) Simplest,
      fastest & least expensive to conduct the review
      - *Cons*: An entity may potentially not be able to apply for its
      label variant (P1v2) in the future if another entity applies for
a label or
      variant which causes it to be visually confusable to P1v2.


   - *Level 2: Existing & applied for Primary labels + ALL Allocatable
   Variants*
      - *Pros:* (a) Relatively manageable pool of labels for comparison,
      except for certain TLDs in Arabic; (b)  Allows entities to
ensure that all
      of their allocatable strings and all other possible future
similar strings
      would be blocked. This creates predictability because all entities have
      their original string and allocatable variants.
      - *Cons*: (a) Certain TLDs in Arabic may have extremely large
      number of allocatable variants, therefore more expensive than Level
      1; (b) 7 scripts in RZ-LGR-5 have allocatable variants


   - *Level 3: Existing & applied for Primary labels + ALL Variants
   (Blocked & Allocatable)*
      - *Pros*: Maximally conservative approach and may reduce the
      possibility of visual confusability among all valid labels in an
      application round.
      - *Cons*: (a) Slowest, most complicated & expensive to conduct the
      review; (b) 21 scripts in RZ-LGR-5 have variants; (c) Certain TLDs in
      Arabic, Cyrillic & Latin may have extremely large number of blocked
      variants; (d) May reject strings due to conflict with blocked
variants that
      will never be delegated.

I previously gave my personal preference to the working group as being
Level 2. However, the more I read through available information and listen
to debate on this issue, the more I lean towards Level 1. Therefore, I
would love to hear what NCSG members think so that the Policy Committee can
agree on an official NCSG preference.
I understand this is a bit of a complex topic, but* if we don't get any
responses by Thursday May 5, we'll assume no official NCSG preference*.

Warmly,
Tomslin
@LinkedIn: https://www.linkedin.com/in/tomslin/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ncsg.is/pipermail/ncsg-pc/attachments/20220430/312b91ca/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: EPDP Team Meeting 30 Slides - E3 E1 E3a.pdf
Type: application/pdf
Size: 965231 bytes
Desc: not available
URL: <http://lists.ncsg.is/pipermail/ncsg-pc/attachments/20220430/312b91ca/attachment.pdf>