Personalization Task Force Teleconference

31 January 2022


becky, CharlesL, janina, JF, Matthew_Atkinson, mike_beganyi

Meeting minutes

Content Module Implementations Status (Follow-up on i18n issue #144)

Matthew_Atkinson: we had an extension but it doesn't work with data- attributes nor all the symbol concepts in the recipe example for #144. we had to modify the extension to work with this example created, so we can show i18n WG and hopefully we can get this resolved.
… , showing a demo
… , showing symbols both LtoR in English and RtoL in Arabic.

JF: ?EA Draffinis fluent in symbol languages
… from the EU

Matthew_Atkinson: also check with Steve Lee

JF: I was wondering about the spacing

Janina: not sure if we need to deal with that for issue 144, but eventually yes we will need to look into that.

Matthew_Atkinson: we just want to make sure this is acceptable to then present to the i18n WG for issue 144

Charles: lets look at two examples RtoL and LtoR beside each other to make sure the symbols do reverse correctly.

Matthew_Atkinson: seems there might be differences in grammar which may be causing the differences in symbols. Need to check with Lion el.

JF: for issue 144 the translation happened and then the symbols were added. We want to make sure they are clear on what we have done.

Matthew_Atkinson: Symbols here I Don't know if they are to flip in a RtoL language. looks like they have, but are they to flip as well? Need to send to Lionel.

JF: I would question why, maybe flipping the symbol we may be changing the meaning.

Janina: we can speculate but lets check with experts first.

Matthew_Atkinson: We can easily add code to flip symbols but we will find out. Ask to discuss on list if we are ready and send it to i18N chairs as well.
… , Lisa said it was ok to share with Sharon and Lionel.
… , I will find some way to send to Lisa if she was unable to get the files.

Sharon: we have a plan for 144

Defining short names for the modules

<sharon> https://github.com/w3c/personalization-semantics/wiki/Personalization-module-renaming-discussion

Sharon: I created a wiki page with all the suggestions in this wiki page. I pulled all the comments as well for each variation we suggested.

Becky: I added some

Janina: I couldn't figure out how to edit or add anything. I was logged in.

Becky: once I was logged in I could edit.

janina: same H3 headings were all the same text for the headings

Becky: I just added some additional options, looking at a Thesaurus lookup

Sharon: we had so many different flavors John recommended putting it in a wiki, and to add comments to try to come to a consensus. It will have a number of updates to make once we make a decision.

Becky: I am happy with Personalization, Adaptive, individualized, transformation, etc.

janina: I would like to keep some form of "personal"… if we drop that we loose years of association with this group. I do like adaptation and to the left of the colon so Personalization / Adaptation, I did want to get away from transformation. Having issues trying to edit the wiki page.
… , might be a bug to file or training. happy to jump on a zoom call to figure it out.

<Matthew_Atkinson> +1 to janina above

<Matthew_Atkinson> This is where JF is searching for specs with "Module": https://www.w3.org/TR/

JF: Do we even use "Module" when searching for CSS in W3C about 50% use Module vs. not using it. So do we really need to use Module specifically.

janina: I think we decided that specifically a couple weeks ago.

Sharon: Yes that has been decided to include "module"
… , should we put a survey to the list and should we narrow this down

janina: yes narrow it down, anyone opposed to "adaptation" be included. I would proposed "personalized adaptation: <name> module

<Matthew_Atkinson> +1 to 'personal*' and 'adaptation' being to the left of the colon

sharon: I would like "personal adaptation: <name> module
… , clarification, I like both personal or personalized

<JF> -1 to adaptation, as many of the attributes do NOT "adapt" anything

JF: we are providing hooks to do adaptation, but we are not doing it.
… , the over arching goal is to add the semantic information.
… , machine recognizable

mike_beganyi: I do like personalized adaptation does seem pretty clear.

<JF> attribute: helptype - "Would be used on additional content to indicate the type of extra help provided. helptype can be used on a span, div, link or image that provides or redirects the user to additional help information."

Matthew_Atkinson: I agree with John just said, our spec does not do any adaptation, but we had discussion about the addition of semantics, but the end goal is to get things adapted to meet their personal needs. It doesn't tell how we are going to do it.
… , if you have any examples where this would be incorrect we should look at that. I agree with you John, but I think this is more about the goals than the method.

<JF> attribute: moreinfo - "moreinfo provides users access to additional information about the current content in one of several formats. A personalization agent may add additional explanatory text or provide an action for the user to take to obtain more detailed information about the content."

janina: I agree with John as well, a number of specs do this we are making it possible to adapt content and get predictable results. I think adaptations is appropriate.

JF: I respectfully disagree.
… , our second module, help type we recognize that there are different types. More info as well. if we use adaptation people who aren't aware of the history. I think Personalization: <name> module is sufficient.

Matthew_Atkinson: so that wouldn't be including semantics which could be confused. there is a misconception that with just Personalized is just a persona

<JF> "The messageimportance attribute is used to indicate the priority level of a message . This can be helpful for a person who gets overwhelmed with messages and filter out low messages and concentrate on critical priority messages."

Matthew_Atkinson: , landmarks some UA don't bother with it, only landmarks can be used by AT or browser extensions, could be a counter example.

<JF> Q:

janina: we discussed that the point that word personalization : was confusing people and expectation were all over the map. Personalization applies to everyone, adaptation strongly is connected with accessibility. adding more help or telling someone the different types of help is a form of adaptation. why I feel strongly to keep this.

JF: I strongly disagree. Message important There is no adapting it is a filtering. I can understand the concern about personalization going back to WCAG 2.1. putting that as a requirement was not going to fly. Yes it can be used for navigation but its also a css selector, the personalization here absolutely benefits a11y but others as well.

janina: this was to do a survey since not everyone is here.
… , we should do a survey vs. email for these two options.

personalized adaptation: <name> module

personalization: <name> module

Matthew_Atkinson: wondering about this issue is adaptation needed to counter any misunderstandings. but the driver for that was confusion. Was it confusion within the COGA group?

janina: No that was for using numbering.

Matthew_Atkinson: would be nice to point to something.

janina: came outside of WAI, email, I will look for it. COGA is about the module 1, 2, 3 etc.

<JF> Personalization Semantics: [term] Module

Matthew_Atkinson: we feel we need to do this because of xyz confusion.

JF: if we feel we need to add a qualifier, we are applying semantics.

Personalization Semantics: <name> module

Matthew_Atkinson: this came up before, fear we may be locked into something in the future, content / semantics but we would like to do both. agree with John yes we are adding semantics, personalization on its own also caused issues so thats is why we looked for a qualifier.

janina: Sharons email is going to say are we voting on 3 alternatives or 2.

<JF> For the minutes, I *STRONGLY* object to Adaptation in the titles

Matthew_Atkinson: Lisa also had some concerns.

janina: I think we are down to 2

Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).


Succeeded: s/?? /EA Draffin

No scribenick or scribe found. Guessed: CharlesL

Maybe present: Charles, personalization, Sharon