Skip to end of metadata
Go to start of metadata

You are viewing an old version of this content. View the current version.

Compare with Current View Version History

« Previous Version 5 Current »

The call for the IDNs EPDP String Similarity Review team will take place on Wednesday, 08 June 2022 at 12:00 UTC for 60 minutes.

For other places see: https://tinyurl.com/c4n2ujtw

PROPOSED AGENDA



BACKGROUND DOCUMENTS



PARTICIPATION


Attendance

Apologies:  none

Notes/ Action Items


Action Items:


Action Item 1: Staff to draft a potential recommendation for review by the small group.

Action Item 2: Schedule next small team meeting for 29 June at 12:00 UTC to review draft recommendations.


Notes:

  • On today’s call: updates on the examples discussed on the previous small group call.
  • Slide 22 – Example 6 – Recap from Previous Meeting
  • Slide 23 – Example 6a – Reviewing Label A
  • Slide 24 – Example 6b – Reviewing Label B
  • Slide 25 – Example Scenarios for Example 6
  • Comment: Could you possibly delegate A1 even if it is confusingly similar to a blocked variant of delegated string B1? Considerations: Can the user actually type that blocked variant? Can a user click on a blocked variant?
  • Staff response: Even if it can’t be typed on an Arabic keyboard, it could be that the characters are on keyboards supporting other languages. The confusability could occur for other language speakers. The motivation comes from SSAC’s recommendations where they talk about the security problem of misconnection vs. denial of service. Even though a variant is blocked there is still similarity between these sets. SSAC advises that misconnection is a more significant issue than denial of service.
  • Comment: The similarity panel will likely include language experts and could take examples on a case-by-case basis.
  • Staff response: The final result of the string similarity review is based on the result of the panel work, but the question is the scope of the review that the panel is being asked to cover.
  • Some support expressed for the approach demonstrated in slides 22-25: compare black and green for Label A with black, green, and red on Label B and vice versa.
  • Suggestion to narrow the comparison. Is there a need to compare level 2 for Label A with Level 3 of Label B?
  • Comment: It appears that inclusion of level three will depend on whether both strings are being applied for in the same round or whether one of the labels is already delegated.
  • Staff note: The proposed solution does reduce the complexity on one side, but incorporating Level 3 on the other side does generate a large number of labels in some cases. This is something to note in the final recommendation if the WG goes with the proposed approach.
  • Clarification: If both labels are applied for in the same round, we are still not comparing Level 3 of Label A with Level 3 of Label B and vice versa.
  • Slide 26 – Example 7: Recap from Previous meeting
  • Comment: Since the Chinese characters are ideograms, if you reorder a sentence, you may still be able to read it even if you recognize that it is out of order. It may be harder to recognize when the text is small, for example with text in the browser bar.
  • Slide 28 – Example 7b: Inspecting Label B
  • Slide 29 – Example 7c: Inspecting Label C
  • Slide 30 – Example Scenarios for Example 7
  • After reviewing this example, some support was expressed for sticking with the proposed mixed-level approach discussed in the previous example.
  • Comment: Keep in mind that in addition to the evaluation that the panel completes, there will be a challenge mechanism to review the string similarity panel’s decisions. Objections processes are also available.

ICANN74 Update

  • Staff will provide a 20-minute presentation on the small team’s progress during IDN EPDP session 1 at ICANN74.
  • The slides are organized following the tasks provided into the small team assignment form. The update will be factual and will not include the draft recommendations of the small team.



  • No labels