- From: W3C CCG Chairs <w3c.ccg@gmail.com>
- Date: Mon, 24 May 2021 15:57:05 -0700 (PDT)
Thanks to for scribing this week! The minutes
for this week's CCG Verifiable Credentials for Education Task Force telecon are now available:
https://w3c-ccg.github.io/meetings/2021-05-24-vc-education
Full text of the discussion follows for W3C archival purposes.
Audio from the meeting is available as well (link provided below).
----------------------------------------------------------------
CCG Verifiable Credentials for Education Task Force Telecon Minutes for 2021-05-24
Agenda:
https://lists.w3.org/Archives/Public/public-credentials/2021May/0107.html
Topics:
1. Introductions and Re-introductions
2. Announcements and Reminders
3. Open Badges v3.0
Resolutions:
1. Marty Reed, Phil Long, and Phil Barker to lead breakout
group on multiple issuers
2. Accepted with breakout group
Organizer:
Kim Hamilton Duffy and Kerri Lemoie and Anthony Camilleri
Scribe:
Present:
Kim Hamilton Duffy, Phil Long, Kerri Lemoie, Phil_Barker, Marty
Reed, Keith Kowal, Adam Lemmon, Nate Otto, Brent Shambaugh,
Dmitri Zagidulin, Jim Goodell, John Gasper, Jeff McCullough, Jim
Kelly, Juan Caballero, Simone Ravaoli, Matt Lisle, James
Chartrand, Stuart Freeman, Stuart Sutton, Taylor Kendall, David
Ward, Jarlath O'Carroll, Nick Hathaway, Kayode Ezike, Balázs
Némethi, Victoria Feng, Peter Mackinnon, Andy Miller, Rick
Barfoot
Audio:
https://w3c-ccg.github.io/meetings/2021-05-24/audio.ogg
<kimhd> Okay welcome So today we are talking about.
<kimhd> A new approach forward to through our schema journey,
starting with open badges and carrier walk us through that carry
Illinois and so let's.
<kimhd> Do it note anyone can participate in these calls,
however, all substantive contributors to any cct work items must
be members of the CC EG with full IPR agreement signed here's the
link to join CG.
<kimhd> and know that that's also in the agenda that was sent out
so before joining you should make sure also will it'll prompt you
in fact that you have a w three web account and so that is free
and open to the public.
<kimhd> And texts of the Community wc Community contributor
license agreement is available at this location.
<kimhd> To call notes these Minutes and audio recording of
everything set on this call.
<kimhd> Are archived at the CCG meetings github repo and you can
see the html versions at this length that I pasted just to note
that i'm a bit behind on meeting post.
<kimhd> archive posting so i'm going to try to get back on top of
that, this week.
<kimhd> Okay, during the call if you have something you want to
add just type Q plus in the chat so that we can preserve order.
Topic: Introductions and Re-introductions
<kimhd> Is there anyone that would like to introduce themselves
start with anyone new to the call or anyone who's sort of
undergone a career shift recently that they like to introduce
themselves in a new context.
<joonas_trussmann> hi, i'm Joonas Trussmann.
<joonas_trussmann> And somebody from the Estonian education sorry
a Heidi seen have working on a educational wallet over abolish
limit the amount of moves, we need to just listen in on this call
i'll be solid from now on.
<kimhd> Wonderful thanks for joining.
<kimhd> For reintroductions sorry to put you on the spot Tzviya,
but if you wouldn't mind reintroducing yourself.
<tzviya_siegman> Sure i'm sorry a segment I work for wiley and I
do a little bit of work with wiley's learning management systems
and IMS and on on the advisory board at the wcc and I did some
work with her viable credentials ended up being three.
<kimhd> hey thanks for joining us Tzviya.
<kimhd> Okay, and, just in case for anyone who did not get the
link directly.
Kim Hamilton Duffy: Here is the link to the agenda:
https://lists.w3.org/Archives/Public/public-credentials/2021May/0107.html
Topic: Announcements and Reminders
<phil-t3> This is phil there's a meeting tomorrow in the early
afternoon.
<phil-t3> For the T three innovation network to mark the.
<phil-t3> kickoff of the next round of work from that
organization and that will be tomorrow at.
<phil-t3> 10am PDT that would be 1pm edt.
<phil-t3> And it's the launch of their so called network of
networks for work group for work channels that encompass the
learning and employment record.
<kimhd> But if you fill in for context teacher innovation network
has been helping a lot with incubating standards in this space in
fact they.
<kimhd> were behind the.
<kimhd> Self sovereign learner employment records work group and
paper and then also that lpr rapper standard which.
<kimhd> has basically it's a ll er as verifiable credential so a
lot of things that have informed this working group, in fact, so
that's a good chance to see what the T three innovation network
is up to as it keeps moving forward.
<kimhd> Okay that's all for announcements and reminders I think
so, with that I think we're ready to move forward to the main
event and.
<kimhd> kerri I was planning to just turn over to you if that's
all right to walk through the proposal and I didn't know if you
had anything you wanted to share but otherwise just i'll turn it
over to you.
Topic: Open Badges v3.0
<kerri_lemoie> Sure, morning everybody.
<kerri_lemoie> And I don't have sides to share, I have some
points to talk over, and then a couple of things to show.
<kerri_lemoie> you hear.
<kerri_lemoie> me is here cool.
<kerri_lemoie> Because he has a chart they.
<kerri_lemoie> can show us to so over the past few weeks we have
been discussing.
<kerri_lemoie> know how to essential pillars right we're spending
a pillar in the context of learning a public record.
<kerri_lemoie> And then we started talking about open badges and
what are the potential opportunities for open badges and if you
know if that is the best way best my best opportunity right now.
<kerri_lemoie> And i'm after you having other conversations
outside of this call, and then, with many of you, which by the
way, thank you for for taking time to talk to me about all of
this.
<kerri_lemoie> The most consistent proposal that came across from
everyone is to focus on open badges because it is the simplest
use case it is already in json ld.
<kerri_lemoie> And also that.
<kerri_lemoie> By implementing a full lifecycle I wonder if I
will put into as a badge we can actually see you know, maybe what
happens, what can happen for other.
<kerri_lemoie> Potential other standards, and we want to
implement as verifiable credentials, so the goal would be to sort
of shift the focus to badges are as a native were looking into
and then complete a cycle like from a Co pilot cycles from
issuance replication and with your at least a wallet.
<kerri_lemoie> And then i'm see what we learned from that so that
we can inform how other verifiable credentials and education
could work.
<kerri_lemoie> Now what we what we think we probably have been
struggling with all along, is that we have all these many use
cases and it gets very complex and as we get down in a certain
way, like Oh, what about this, what about this thing.
<kerri_lemoie> And, but if open badges themselves can be very
complicated, even just either on simple.
<kerri_lemoie> approach it, and I say simple release to the
madness and some other potential opportunities in doing this
could be that we look at open badges.
<kerri_lemoie> As a native verifiable financial so instead of
looking at other options that we looked at last week, the week
before instead.
<kerri_lemoie> We just say hey what is an open badge as a
verifiable financial and just go for it.
<kerri_lemoie> And then, while doing now, we have other
opportunities for the matches, we could do things differently,
like distinguish the difference between a crater and an issuer.
<kerri_lemoie> We could add a credential type and we could also
that that property is from schema.org to to enhance the
properties that could be in a batch.
<kerri_lemoie> So I wanted to start there with this proposal and
to see what folks think and and see if we can reach a consensus
on nine so, then we can sort of look at.
<kerri_lemoie> You know what direction, this would go in and
maybe it's either a very early example of what a no magic the
quake as a native vc.
<kerri_lemoie> so you have a question, so why don't we start
there oh.
<phil_barker> yeah.
<phil_barker> I think this is a good idea I think it's a good
idea to start with.
<phil_barker> Open badges for the reasons that you laid out and
but what I think we should keep an eye on.
<phil_barker> All of those complicated use cases that whole sort
of mess of things that have been knocking off because of course
we shouldn't let it let them knock us off course.
<phil_barker> But I think we should look at some of the things
that might be problematic with open badges before we go into too
much depth of how you model, an open badges a native verifiable
credential shirt so.
<phil_barker> We should.
<phil_barker> try and get Anthony back on one of these columns to
talk about his use cases of multiple issues and how you can
address that with.
<phil_barker> The open badges we should look at the different
types of things that you might want to credit, you know the.
<phil_barker> The three top level concepts, is it three anyway,
the top level concepts that we worked through over the last few
weeks, and I think we should look at those before we go into too
much depth on you know small little england's to be open badges
models.
<kerri_lemoie> yeah absolutely.
<kerri_lemoie> I don't want to abandon the more complex use cases
I just don't want to get bogged down with them.
<kerri_lemoie> I think that, like, for instance, open badges they
have one issue or verifiable potential has one issue or could
they be multiple issues that are like combined into one issue I
mean I might happen to go badges now but.
<kerri_lemoie> we're still there's only one issuer and the
credential, and so there are some similarities already between
open badges assertion and a verifiable financial assertion, I
also think that this many ways can inform how we can approach CLS
as verifiable credentials.
<kerri_lemoie> i'm looking at the chat here.
<kerri_lemoie> A year later.
<kerri_lemoie> yeah i'm sorry i'm excited to see phil on.
<phil-t3> nate's up i'm just saying.
<kerri_lemoie> Oh Nice.
<kerri_lemoie> Okay.
<ottomomy> yeah I was just saying a multiple assures sounds like
a verifiable credential level concern, and so we can elevate that
back up to the full Community group.
<ottomomy> This is not in the scope of education to offer really
like a what is the conical.
<ottomomy> verifiable credentials way to do something concepts
such as multiple issuers, we can describe other more limited
scope problems here, probably and then eventually model them like
an issuer that has a relationship to a parent organization
something like that.
<ottomomy> And I don't think that wouldn't necessarily block
consideration of any particular modeling.
<ottomomy> A path forward but yeah we can consider some more
advanced cases I think that allowing ourselves to be bogged down
with and stop ourselves from pursuing a model would be an Anti
pattern.
<kerri_lemoie> Thanks, I put one of the give every place in the
chat.
<kerri_lemoie> Yes, totally agree with that i'm Kimberly you had
a question about full issuers.
<kimhd> yeah I wanted to also talk to multiple issues and
carriers my co chair, I encourage you to shut down conversation
on me if I get too long winded so.
<kimhd> My my thought is absolutely what made is saying, and I
think we have discussed multiple issuers already in the document
the vc ED models document.
<kimhd> we've talked about an approach to multiple issuers at the
vc at the verifiable credential level, I think the one other
thing that.
<kimhd> You know, we could consider is just clarity around what
we mean by issuer because I think frequently when we get bogged
down in multiple issues we're talking about all different kind
of.
<kimhd> sort of I don't know certification bodies or something
that we don't actually mean by a verifiable credential issuer, so
I would actually I don't think we need to block on that, I think.
<kimhd> Maybe that's a good idea for a focus group if okay if
maybe here's a proposal if if people think that that's very
important or they don't and they are sort of.
<kimhd> You know I think in my mind, I see that it is a.
<kimhd> question that doesn't need to bog down further
discussion, however, I get that not everyone would accept that
so.
<kimhd> If people would like to move forward on that, I think
that might make sense to do a breakout group to come back with a
proposal and clarity, otherwise that's something that i'd rather
just leave that be, for now, because we've given a lot of
attention to it.
<kerri_lemoie> Like the idea of a breakout group is there anybody
on the call right now, that would be interested in leading a
breakout group to discuss multiple issues they would like to
bring it back to us.
<kerri_lemoie> In a future call.
<phil-t3> i'd be happy to be on that group i'm not sure I have
bandwidth to lead it leading it also means summarizing all the
work.
<phil_barker> yep similar here it's Anthony who's raised this as
an issue a couple of times and I don't think we should discuss it
now, I think the opposite, I think we should get something solid
as a proposal for dealing with.
<phil_barker> Open badges and then see how we can address
multiple issues within that framework, and I think it does need,
we do need to go back to Anthony to get better a fuller
description of him some concrete examples of where multiple
issues and needed.
<kerri_lemoie> On this call right now.
<kimhd> No he's not been for a while, so I think that you know
it's it's on him to sort of follow the output of the group, and
you know, make sure it's abby so I don't think that he needs to
be a gatekeeper here.
<kerri_lemoie> and
<phil-t3> Marty I don't want to put you on the spot, but I know
that you're dealt with this problem and.
<marty_reed> You certainly would be a voice that would be.
<phil-t3> My question is, is there a way that the definition of
an issue or can be sufficiently bound, that the issues,
corresponding to those that have what seemed to be more of an
administrative political need to delegate legal responsibility in
a chain.
<phil-t3> Is it possible to deal with those outside of the scope
of the actual code that's that's defining the issue or of the
particular badge in question.
<kerri_lemoie> I think bill, that is all of those things really
why we want to set it aside as a separate conversations to maybe
we put this in our back pocket, for now, and then we revisit it,
because if folks think it's important we should address it, but
also.
<kerri_lemoie> I don't think it is the simplest use case that we
can tackle first right at the simplest use case is that we have.
<kerri_lemoie> An issue shooting a badge and recipient, and you
know it can go to a wallet and then they the badge gets sent from
the wallet to a presentation to verify.
<kerri_lemoie> It we really like to just tackle well we don't
have it's funny you on this call last week and.
<kerri_lemoie> It was like all these complex cases of like
wallets and how are they all going to use the word he had to
prove that even wonder if I will financial education can be moved
around in this way.
<kerri_lemoie> That we all feel very good about and so why don't
we just happen like that really easy space.
<kerri_lemoie> And then, yes, come back in and take care of this
complex use cases and I think we're talking pretty fast right, I
think we would like to get through the simple use case like until
the early fall and then really see what we've done with this and
then move on from there.
<marty_reed> So i'll throw my hat in the ring for the breakout
group I was debating on if I had bandwidth, but this is very
important topic so all right don't.
<phil-t3> count on it to then Marty help you and phil Parker,
would you be willing.
<marty_reed> Those are those that are being me.
<marty_reed> to join that breakout.
<kerri_lemoie> As great money Thank you and then maybe we paid a
report, but could you let us know, maybe like after you on mute
once when you'd like to sort of come back to us with it and we'll
make sure that we include that.
<kerri_lemoie> On one of the calls when we will feature what
you're doing.
<kerri_lemoie> Thank you.
<kerri_lemoie> Okay, so I guess like I said open this up a little
bit more of a discussion to see there's any other concerns about
taking this approach.
<kerri_lemoie> And not him i'd like to make a proposal that we
see we have consensus on this approach.
<kerri_lemoie> Here is any objections.
<kimhd> In the chat type well last one, if you want to endorse
minus one with the comment if you're actually just minus one and
then we'll call on you, if you have any objections.
<kimhd> To terry's proposal so that proposal just let me write
that down so proposal was move forward with open badges as
actually can you say that again.
<kerri_lemoie> Oh yes, the proposal is to move forward with a
simple example a full cycle education example of open badges as a
verifiable financial has that.
<kimhd> Okay, that looks good.
<kimhd> So mark, this is a resolution.
<kimhd> Here accepted that.
<kimhd> The exception being the breakout group i'm imagining
Marty.
<kimhd> yeah.
<kimhd> I have that noted to.
<kimhd> Okay excellent he just joined by the way.
<kimhd> Alright, so I guess, we can keep moving forward.
<kimhd> kerri what was the sort of next part after that you
mentioned that perhaps nate had something that he wanted to
cover.
<kerri_lemoie> yeah so i'm going to be meeting some Venus child,
so I am going to screen share it then i'm going to hand it over
to need to sort of explain it through.
<kerri_lemoie> And then, what I have is a very, very rough
example of what an open badge as a verified look look look look
like as a beta verify looking at.
<kerri_lemoie> i'm going to share my screen right now.
<kerri_lemoie> All right, can you see it.
<kimhd> Yes.
<kerri_lemoie> Okay, did you explain your chart for.
<ottomomy> Sure, so we've talked about the idea of open badges a
number of times before, and you know, the idea that had been
implemented by open badges it's a generic concept that we've
talked about a lot with the idea of a defined.
<ottomomy> achievement that is a bundle of information provided
by a specific issue or about a learning opportunity, maybe, an
assessment.
<ottomomy> An image that represents the the achievement, a
description and specific criteria of what it means for an
individual to have earned it.
<ottomomy> And that, in order to fit this already json ld
structure as it's been implemented an open madness and other you
know linked data structures, like, for example with schema.org
education occupational credential.
<ottomomy> into a verifiable graduate, we need to find one that
particular essential claim, which is what does the credential
subjects relationship to this concept of the defined the
achievement that they have earned look look like, and so in this
diagram that.
<ottomomy> This is a verifiable credential up at the top, and in
the box here between the credential subject has been identified
with a did and that.
<ottomomy> definition of the achievement, there is a you know,
has credential property of some sort and schema.org house
credential.
<ottomomy> That can be can be used, you know our decisions here
in this group are almost limited to what's has credential look
like and then, if we look at the other types of data that are
present in assertions of this type, you know open badges being an
example.
<ottomomy> There is a concept of evidence there's the concept of
you know, an issue date an identifier for the the overall
assertion itself.
<ottomomy> there's the concept of an issuer, that is, the issuer
of that assertion and so it's a matter.
<ottomomy> of just sorting all these different properties that
exist within our conceptual model into where they go in the
verifiable credentials model, a bunch of them fit right at the
root of the verifiable credential notably issuer.
<ottomomy> The issue dates.
<ottomomy> The overall ID of the credential and some form of
proof.
<ottomomy> Those are all concepts that were in open badges two
point O in the assertion model those can live at the verifiable
credential level, and then the other set of properties that
you're filtering out are at the.
<ottomomy> claim level is what are you saying about the
recipient, and the only tricky one I think is is evidence that's
idea that there is supporting evidence of portfolio item it's all
modeled on the schema that are creative work class.
<ottomomy> there's a piece of work or multiple pieces of work
that represent why the recipient has earned the criteria of the
grandchild, and that makes perfect sense as a.
<ottomomy> claim of some sort about the recipient, so the idea
here is you're saying the recipient has a credential and they
have evidence as sort of parallel claims that you're making in
this one bundle or sorry I say, has an achievement.
<ottomomy> Has a credential property a couple other things on
this chart.
<ottomomy> In there's a good handful of other variants on this
that we could talk through but all i'll stop here and and ask for
questions is that there's this idea of a result against a skill.
<ottomomy> So there's a concept from that we've talked about here
as well, the open the skills assertion, the idea that many
different issuers might.
<ottomomy> want to make a claim about the achievement of a
particular result that might not have been defined by the issuer
but let's there's some other examples that will talk to that a
little bit more precisely so i'll stop here and turn it back over
to okay.
<kerri_lemoie> sorry about that yeah we.
<kerri_lemoie> will leave this up for a segment, while we see if
anybody has any further questions or wants to discuss this at
all.
<kerri_lemoie> anyone have any thoughts on this so far.
<tzviya_siegman> Can just ask for a little bit more clarification
about the evidence and creative work.
<ottomomy> Yes.
<tzviya_siegman> So what that is.
<ottomomy> Sure, so, for example in open badges The idea is that
i'm going to award an assertion to carry and i'm going to.
<ottomomy> The main essential part of that assertion is just
which badges it for which badge definition and then there is a
section that allows evidence to be.
<ottomomy> claimed as well, so I can say, as the issuer Okay, I
have seen good evidence that that kerri is is deserving of this
badge and so i'm going to describe a little bit of that evidence
here in.
<ottomomy> In the assertion and the evidence class that as its
defined by open badges it's based on schema.org creative work in
that it has a name description a URL and a lot, you know just
basically lets me describe what evidence i've seen that.
<ottomomy> convinced me that very deserved will have a particular
badge, and this is included in the assertion and kerri can carry
it with her.
<ottomomy> And Whenever she shows off the air and assertion that
you've got links to the evidence, are there so that the
verifiers.
<ottomomy> can take a look at what it is that she's done it's not
embedded files, although some people have come up with clever
ways to be 64 encoded stuff and train jam entire pdfs inside a.
<ottomomy> badge json but it's just the idea is it's a
description of a piece of portfolio item, essentially in terms of
the URL where you might find it and or descriptive metadata like
a name and a narrative.
<tzviya_siegman> The reason I was particularly asking is that I,
one of my job's is chairing the publishing working group and
there's a lot that when you get into things like bibliographic
metadata it gets complicated and I just wanted to make sure we
weren't going down that path.
<ottomomy> Evidence identified by a URL link here, in most cases,
and sometimes just description.
<> Will.
<kerri_lemoie> Have a question.
<phil-t3> Both of us to.
<> Are in the queue.
<phil-t3> But.
<phil-t3> Let me yes i'm in the queue.
<phil-t3> And I just wanted to clarify it helps at least when I
describe it to folks typically describe the evidence as as neat
correctly noted, it is a your your eye that points to something.
<phil-t3> That may be a directory that may be a file that may be,
whatever whatever the the target happens to be, but the point of
it the intention of it is to provide the person who is reviewing
an assertion or a claim, not only the opportunity to know how the
the claim was was.
<phil-t3> determined that is associated with some kind of rubric,
which is also usually a part of a badge but also the methods of
assessing how that rubric led to whatever the level of
performance is.
<phil-t3> that's being asserted, and then you have the person's
own work for independent set of eyes to say I see here's pieces
of evidence that correspond to doing very well on these math
using these methods of assessment to determine the assertion this
person knows how to ride a bicycle.
<> Thanks.
<kerri_lemoie> For barking.
<phil_barker> There can only be one achievement recipient right,
otherwise you don't know which evidence goes with Richard
treatment correct yes.
<phil_barker> That remains true if you take this data, out of the
sort of the json ld rapper and put it into a graph of some so
you've got to keep them associated to sort of single atomic yes
or no, you know isolated graphs you can merge the graph can't
merge this graph with a graph for another credential.
<phil-t3> Just to clarify So if you there can be multiple at the
pieces of work that correspond to demonstrating that one
achievement right, yes, yes, yes Okay, if you wanted.
<phil_barker> To have to correspond to the one on one.
<phil-t3> And where there's a circumstance, one could credibly
make that it's an exam, it is a contributor to more than one
achievements, can it be.
<phil-t3> reproduced in each of them.
<phil_barker> I think it's more of a problem if you've got more
than one achievement right different evidence yeah yeah that's I
seen a snake wants to respond, sir.
<ottomomy> yeah only to say exactly right.
<ottomomy> If you are running a system that is doing some json ld
processing on the graph and sort of constructing a graph of
claims about the.
<ottomomy> recipient here that's identified by I did, one of the
things that you might do is bundle all of the House credential
claims together and say this recipient has 45.
<ottomomy> credential has credential defined credential
achievements and then you might bundle all the evidence claims
together, you say this person has 30 portfolio items that have
been declared about them as as evidence for.
<ottomomy> As credential claims, but in order to be able to
figure out which goes with which you would have to maintain the
original pairings you know the original.
<ottomomy> unit here that's defined on the screen this particular
set of claims pairing the has credential claim and the evidence
claim and all those types of processing are completely reasonable
to do.
<ottomomy> And if an inspector does want to break them out, they
would have to understand that when you're defining a credential
that has multiple properties like this sometimes it's useful to
be able to track.
<ottomomy> Those things together as a single credential with
different claims that go together and sometimes it might be
useful to track just.
<ottomomy> This one particular claim.
<ottomomy> across.
<phil_barker> Multiple credentials So yes.
<ottomomy> What are all that has potential relationships this
subject as.
<kerri_lemoie> I said I can has a chance to.
<kerri_lemoie> respond to that before she's in the queue please.
<kimhd> Let can go ahead because mine is it a separate thread.
<kerri_lemoie> Okay, thanks for the heads over.
<phil_barker> The children to serve the achievement is
essentially a bunch of classes, no.
<ottomomy> that's right yeah okay.
<phil_barker> yeah I was.
<phil_barker> just trying to think if there's anything else you
can attach the you know that the evidence and the result to that
doesn't lead to that sort of.
<phil_barker> dissociation between.
<phil_barker> What they've done and.
<phil_barker> What the evidence is that provided for doing it.
<ottomomy> kerri after you kim's Fred maybe we go look at the
skill assertion example in the top right.
<ottomomy> To answer that question.
<> Sir.
<kimhd> So mine is is probably a little basic because I was
dealing with a pug freak out momentarily and I may have missed
some of nate's framing.
<kimhd> So I think I was getting hung up on things like I was
expecting to see terms like badge class and everything, and I was
wondering was this is this meant is sort of a conceptual diagram
like because I guess what I was wondering is like.
<kimhd> Taking that diagram at face value, it seemed.
<kimhd> Pretty disruptive on top of the current open badges
model, so I was just wondering if it was you know in in maybe we
can get away with that, but I was just wondering about the
framing of it which you probably mentioned already and I.
<ottomomy> didn't mention that but that's a very good question so
i've used in this diagram of use generic terms such as
achievement that kind of speak to if you actually go back to the
previous one, that we were looking at in the middle.
<ottomomy> The.
<ottomomy> So i've used generic terms like achievement, instead
of open badges specific terms like badge class it's also worth
noting that car, which is another IMS spec.
<ottomomy> adopted the open badges assertion badge class and
profile classes and they renamed batch class to achievement
there, so this is actually a CSR term to see achievement.
<ottomomy> Here, just like actually this example contains
something called a result as a result description which also our
terms used in I msc lr but.
<ottomomy> As we get more specific I don't think i'm proposing
any necessary, but any name change to those classes necessarily
occur, but this is not an IMS working group here and the vc EDU
Paul, this is a.
<ottomomy> w three see space, and so I thought it was probably
better to use some generic terms to talk about the generic
concepts that were wrestling.
<kimhd> Thanks nate that looks great.
<kerri_lemoie> And then, before I get to what I had before need
to enter your address your saying about the skill skill session
here.
<ottomomy> Sure yeah another concept that we've talked about for
the last year and a half or so in this group is the idea of an
open the skill assertion, where there is a defined.
<ottomomy> skill of some sort or competency, perhaps a learning
target that exists in the world it's been published, with a
unique URL the open skills network is doing.
<ottomomy> defining a concept called rst or rich skills
descriptor which is generic pattern that could be implemented in
several different link data friendly.
<ottomomy> standards that just means that there is a skill
definition or competency definition, it has a unique URL and when
you fetch that URL, it is possible to get.
<ottomomy> machine readable metadata about the skill, so we can
treat it as Oh, there is a skill definition authorship of that
skill definition is an important concept that skill has been
defined by a particular author.
<ottomomy> Once that skills out there, you could imagine, maybe a
industry organization professional association or something like
that might be one organization that would define skills
educational institutions are another common one state.
<ottomomy> boards of education play in that space as well we've
talked a lot about the definition of.
<ottomomy> Skills and competencies and frameworks of those
things, but now we can talk about the assertion of what it looks
like when a particular individual has that skill.
<ottomomy> has achieved it at some level, and I have what would
that would look like in a verifiable credential, and so this is
pretty similar and.
<ottomomy> compatible graph with the other one that there's a
verifiable credential with all the properties there and it a
credential subject identified by a did.
<ottomomy> But here, in this case, we can make a slightly
different claim about this skill, by saying that the.
<ottomomy> holder here has achieved a particular result against a
skill what I mean we could have constructed a direct relation to
skill like has scale or something like that, but I think this
intermediate model.
<ottomomy> places a gives us several other use cases that we've
talked about within the scale assertion model, such as the
ability to maybe express at which level.
<ottomomy> In the skill did the individual perform and is my do I
have a degree of confidence about that.
<ottomomy> It allows the author, the issuer of this credential to
make some some more detailed claims about the achievement.
<ottomomy> And the other cool thing about doing it this way is
this fits really nicely with a achievement.
<ottomomy> claim as well, so Kim if you could look at their Kim
kerri if you could scroll to the left one slide and look at the
combination graph where there is both a has credential
relationship and the skill result relationship in the same.
<ottomomy> chart these things can can go together sometimes the
skill that you're.
<ottomomy> Recognizing independently is also something that the
achievement has aligned to so you could say this achievement
aligns to this skill and then you could say.
<ottomomy> And there is a specific result about that skill that
i'd like to record for this individual so that I can express
concepts such as level and confidence.
<ottomomy> So we've got a couple couple different options here,
but overall there's two main use cases being represented in.
<ottomomy> These graphs here's how with a verifiable credential
we could express that individuals have.
<ottomomy> A particular credential but they've met the criteria
of a defined achievement, where the issuer of the verifiable
credential is the assure or is clearly authorized by the issuer
of the.
<ottomomy> defined achievement to make that claim, and then the
separate type of claim, where we are saying that.
<ottomomy> A individual has a particular skill or competency and
the skill or competency may be defined by a different
organization than the issuer of the credential itself is making
this claim today.
<kerri_lemoie> Okay, thank you i'm Anthony you have a question.
<anthony_camilleri> Oh yes, oh I like this conceptual model and
what i'm wondering is i'm looking at people achieved result and
the evidence arrows.
<anthony_camilleri> And i'm wondering why these not extend from
achievement, rather than from the ID recipient.
<anthony_camilleri> For example, the evidence is not evidence of
your identity it's evidence of your achievement so conceptually
speaking, it would seem that it belongs nested under treatment,
rather than nested under recipients.
<ottomomy> Well what's important to understand about this
achievement model is that this is a achievement definition that
might apply to any learner This is like the bachelor's degree in
science it's not nate's bachelor's degree in like not to be yes,
I have a BA.
<ottomomy> it's not nate's degree, this is me this is the
credential that would be awarded to any.
<ottomomy> person, so the claim that we're actually making for
the sort of core claim of the graph here is that this particular
holder has achieved the criteria of a generic achievement that
many holders.
<ottomomy> have achieved, and then the claim that we're making
with a particular result is that this holder has a particular you
know, has achieved a particular result against some kind of
target maybe a skill.
<ottomomy> And then, a evidence claim that we're making is that
this particular holder has a piece of portfolio.
<ottomomy> Item another way to do this chart absolutely would be
to have the assertion model be kind of an intermediate thing in
the graph here where we're saying that.
<ottomomy> The central claim would be this recipient has an
assertion.
<ottomomy> And then that assertion has a number of relationships
to other things that assertion is associated with evidence that
assertion is associated with skill claims and with a particular
defined achievement.
<ottomomy> I think that this structure, without that intermediate
model is cleaner and allows us to do some of the cool.
<ottomomy> You know, breaking these things out as direct claims
again for the recipient, but other people might feel that a sort
of has assertion model would be a better way to go, and this
group is a great place to discuss that.
<phil_barker> Can I jump in.
<> Leslie.
<phil_barker> And I just want to confirm what I think I heard you
say that the skill result is specific to the subject to the
person it's what that person achieved that might make.
<ottomomy> that's what i'm proposing is that we say a person
achieves a skill result and that's you know you can say that they
have a skill through that result.
<phil_barker> So why not attached the evidence to the skin
result.
<phil_barker> Because it's evidence that they have that result.
<ottomomy> Not doing it this way is just a little bit more
granular it allows you to the evidence is associated with both of
the claims in this credential the chief resolve and the House
credential claim.
<phil_barker> I mean, I guess, I guess, we could have had it.
<ottomomy> Well, if you want to issue two different credentials
to in order to make that more specific go ahead yeah like I said
on that chart on the right.
<ottomomy> There is another option for this position doesn't have
that has potential in it, and so, if you want to make that.
<ottomomy> claim the claim that there is a skill result that has
evidence cool just do this chart and then, if you want to make
the claim that the person has defined achievement and there's
evidence separate make that claim as a separate potential as.
<phil_barker> Well, what.
<phil_barker> What i'd like is for the claims for the evidence in
the claims not to fall apart, when you aggregate this data so Kim
wants to interrupt a letter.
<kimhd> Oh thanks yeah I think that's a good conversation, but I
was also thinking that sounded like something we could.
<kimhd> It sounds like a revision of the model that maybe we
could come back to later, I think that there's it sounds like
there's alignment that this is a promising approach but.
<kimhd> We have feedback to it, and so I would propose that maybe
we come back to it because I thought that maybe I saw some other
things like maybe I thought I saw some samples as well, so I
thought that might be helpful to look look at that too.
<kerri_lemoie> I mean, I do think this is what we should connect
these conversations, and this is why.
<kerri_lemoie> Why, even the most simple digital credentials
specs has all had a conversation associated with it.
<kimhd> So Nick fury actually like if if you'd rather I was just
sort of trying to sort of keep track of time and everything So if
you think that's the priority right now.
<kimhd> I just thought I saw a few other things that might be
useful for the well rounded perspective, but then otherwise i'll
let you to sort of guide this.
<kerri_lemoie> No, no, I think you're right, I think we don't
want to get into too many details right now and, in fact.
<kerri_lemoie> The one thing I have to show you is not we really
don't want to get you tell some because I literally threw this
together, like right before the call.
<kerri_lemoie> Like I had some of these, but I wanted to get a
get together an example to show all of you, but i'm sure it is
not correct.
<kerri_lemoie> And I don't want us to like spin on this at all, I
just wanted to sort of give you a sense of what it could look
like for an open badge to be a verifiable credential.
<kerri_lemoie> it's actually not so different from what Neil has
been proposing and chemists and proposing in the past, and this
was sort of emerged of the many options we discuss, which was to.
<kerri_lemoie> You to have the verifiable potential right to be
the the essentially replaced the assertion of an open batch.
<kerri_lemoie> Where we would have the issuer be part of the
verifiable credential and the issuance date, and then we have the
parents are subject and I don't have any of this matches what on
the chart right now, so please keep that in mind, but we have
like an identity for the recipient.
<kerri_lemoie> We may also want to consider like what do we do
with an email right approach and badges and it's been used for so
long, so that's, not even in here that discussion, yet not that
it's been left out on purpose just said, I hadn't put it in yet.
<kerri_lemoie> But this example that has crenshaw does have a
badge and that badge has an ID and it has a name and a creator.
<kerri_lemoie> I was messing around with this concept of you
know, the creator of the badge classes some allowed issuers just
just playing around with what that could look like very lightly.
<kerri_lemoie> i'm a bad image description care what I sort of
like about all of this, though, which you don't see is like a big
batch or you don't see an image anywhere like a required admin
chicken.
<kerri_lemoie> I mean, I know for badge always has required this
image and that data was baked into the image.
<kerri_lemoie> This could we could change this entirely with
badges as vcs we can still include an image, the image doesn't
need to be required, which will be pretty huge right i've been a
different approach, but also give you the saver French.
<kerri_lemoie> So people could still have an image that they
wanted to take a still baked into the image that they wanted to
but it's not required.
<kerri_lemoie> These are just some thoughts here some
opportunities that we have to discuss and then here's where
evidences and I think I typed it in where I thought it should be,
which is under has potential that you can always look at that
later.
<kerri_lemoie> Where it's even richer than just what's in
evidence badges now but it's like includes creative work where we
can do more than just what we had in evidence, so this also gives
us an opportunity to introduce new.
<kerri_lemoie> Additional schema to have met is that we didn't
have poor terms of making it more rich set.
<kerri_lemoie> Okay i'm going to pause there was a whole bunch of
things going on in the chat so take a look at this.
<kerri_lemoie> See.
<kimhd> i'm late.
<kimhd> I was following kerri kerri maybe i'll TEE up certain
people whose comments like it might be good to say them out loud
i'm.
<phil-t3> arrogant.
<kimhd> He won on multiple trust anchors.
<phil-t3> yeah i'm just saying that this is actually I think as
much a digital identity group discussion and getting groups
getting their their perspective on ways to handle that
particular.
<phil-t3> problem might be useful on there, I know there are
people working on this, who are coming up with ways of
associating different levels of confidence to.
<phil-t3> To representations of identity in a key chain like
fashion and i'd wanted to replicate that work because it's it's
it's challenging by itself just suggested we we reach out to some
of those and get some input.
<kerri_lemoie> yeah that sounds good yep.
<kimhd> So I want to keep in mind first Kerry said this was she
did this right before the meeting so with with the understanding
that.
<kimhd> You know we'll probably go through some revisions on this
Keith did have a question around should this example, include the
assertion or public page URL I think that's interesting because
this gets to some comment that.
<kimhd> Naming carry both in mentioning around links versus data
in line badges so i'm wondering if Keith or sorry if nate or
carry have anything to add on that sort of URLs versus in line
data.
<kerri_lemoie> Well it's funny because in all it needs to be
talked about in some detail a little bit anyway, and I would say
a lot.
<kerri_lemoie> I had actually even left the ID out of the badge
when I first was talking tonight about this and you know, there
are good reasons why you would want to have like a very specific
canonical URL for a badge meaning the batch class in terms of the
assertion having an ID I don't know.
<kerri_lemoie> If it's going to be a verifiable credential do we
want to have hosted versions of verifiable credential and I think
that is another another conversation to have i'm there are
reasons why you would want to have a simple hosted badge versus a
badge as a verifiable credential.
<kerri_lemoie> So I think that's that's a discussion, we need to
have that that anything that's my answer to that right now.
<kerri_lemoie> may have a year.
<ottomomy> So I think the reuse ability of a defined achievement
is a critical capability that that open badges has.
<ottomomy> In particular, as a standard that implements this
generic concept has done really well, I mean it didn't it wasn't
always that way, so there was 0.5 badges that did not have unique
ids for the defined achievement class.
<ottomomy> Batch class and then that was added in 1.0 as a method
of sort of normalizing the data and I don't know that the
architects who worked on one point O necessarily.
<ottomomy> really thought through a lot of the consequences of
that change.
<ottomomy> They were kind of like it's more efficient if we only
publish this data once and, but the fact that it's uniquely
identified gives it a lot of power, and so the platform that we
build badger.
<ottomomy> is an example of why that's important which we have a
concept called learning pathways in our product, where you can
require that a learner completes three of these five badges in
order to.
<ottomomy> be measured as complete on a particular step of a
learning pathway, and without a unique identifier for the badge
class.
<ottomomy> it's not possible or it's not practical to say that
the learner must complete these you know, three of these specific
five badges there's just nothing to attach that that.
<ottomomy> request to that rule to and so unique identification
of a defined achievement, I think, is a really important part of
defining achievement achievement.
<ottomomy> Open badges or Point five are really good sort of
storytelling tool they show the data, you know you can render it
into html templates if you want for human eyeballs but they don't
do very well for building.
<ottomomy> Business logic around like the meaning of digital
credential having unique identification, for that is essential
and we can go even into further rabbit holes around like are
there, different versions of the data with with this ID.
<ottomomy> etc, but we can grow in that capability i'd rather not
sure that capability away as to the key question around the idea
of the assertion itself and whether that idea would be like an
http.
<ottomomy> identifier, that would be essentially a public page
for this credential I think that is a implementation option that.
<ottomomy> would definitely be open to organisations and how the
privacy and access control would be managed on such a page up to
the employment do is to determine and maybe some spec office to
recommend.
<kerri_lemoie> Kim you had a question or statement and not my
question.
<kimhd> So I I wrote down this topic to come back to the idea of
you know links for the badge class versus links for the assertion
is something that has come up a lot in terms of.
<kimhd> You know the assertion can contain pii or maybe you want
it to be able to have something, and so I think.
<kimhd> That the badge class is frequently something that you
want to maybe attach more data to over time, for example, and you
want people to be able to refer to, so I think like that's a
really good.
<kimhd> discussion, we can come back to introduce making badges
more badges via this vc alignment kind of more usable and
flexible for different kinds of.
<kimhd> You know, sensitive data kinds of claims, so I think the
other one that's interesting, you mentioned the fact that they
bad the baked in this is not.
<kimhd> Here, but that also seems like an interesting thing to
come back to so i've added this you know just in terms of how we
deal with the idea of you know.
<kimhd> aligning the image image with the tamper evident payload
and all the many problems that come up with that, so we have an
opportunity to if we want revisit baking and sort of how that
comes about, but so that's all things that I think of is
appropriate for down the road discussion so.
<kerri_lemoie> Thank you, he was up next yep.
<keith_kowal> To start, I mean it's just a higher level comment
that the beauty of open badges is that you can use them in
today's world, so I mean.
<keith_kowal> If you go to a pure vc model than you are stuck
with chicken and egg problem so they're not being a verified
ecosystem to.
<keith_kowal> to consume those vcs so by putting in the links to
a public page or to an assertion URL it allows you to you know,
to have a middle ground where you can continue to like share that
link on linkedin or Facebook.
<keith_kowal> And again to what nate said, you know it can be
optional, they just added will be good and examples, just to show
it because I think in the interim phase when there's no relying
party system ecosystem to consume verifiable credentials.
<keith_kowal> To me, this is the one of the great reasons why I
really support putting open badges and vcs because it gives you
something you can actually do with the vc while we're waiting for
ecosystems to step up.
<kerri_lemoie> yeah that is awesome I agree i'm an idiot and I
know we're going to save this for later and I just have one more
other thought on the the assertion URL is that right now.
<kerri_lemoie> We don't really have systems consuming who are
like assertion data either, and all the sharing that can happen
from platforms doesn't require an assertion.
<kerri_lemoie> That this requires a webpage that is hosting the
badge it doesn't require an actual assertion, so I just want to
put a pin in that for future we have that future discussion on
that.
<kerri_lemoie> Because I will forget holly between now and then.
<kerri_lemoie> The yes, thank you for bringing that up and then
because we do want to make sure we're bridging right, we do not
want to like move a lot of agile happening right now, so we want
to make sure that continues and continues to support that
ecosystem that exists.
<kerri_lemoie> Do you see Is there anyone else here in the chat
that has something with the sake that point or any other point.
<kerri_lemoie> Alright, well, I think we already have some topics
for future call skin.
<kerri_lemoie> I don't think I have too much else to share at
this time.
<kimhd> yeah I think that's great Thank you kerri for leading
that and I think that let's see yeah so I don't think that we had
anything else from the chat to add, so we can we're getting close
to time, so we can wrap here.
<kimhd> Thank you everyone and i'm going to try to make some
progress on publishing the backlog of minutes this week.
<kerri_lemoie> yeah Thank you to all of you for these
discussions.
<kimhd> Following are from notes Kim and Kerri took during the
meeting
PROPOSAL: breakout group on multiple issuers
RESOLUTION: Marty Reed, Phil Long, and Phil Barker to lead
breakout group on multiple issuers
PROPOSAL: move forward with full cycle implementation sample of
Open Badges as a VC
RESOLUTION: Accepted with breakout group
Kim Hamilton Duffy: Topics to return to:
<kimhd> 1. link to defined achievement vs link to assertion
<kimhd> 2. baking in v3
<kimhd> 3. bridging with existing deployments vs current proposed
design
<kerri_lemoie> 4. Review, present, categorize existing use cases
related to Open Badges
Kerri Lemoie: 5. Open Badges VC Instances vs Assertions:
Comparison of Use Cases
Kerri Lemoie: 6. Open Badges VC Images: Who Needs them?
Kerri Lemoie: 7. Open Badges VC: credentialSubject recipient
Kerri Lemoie: 8. Open Badges VC: hasCredential (badge: name,
description, criteria, image, alignment)
Kerri Lemoie: 9. Open Badges VC: creator & validated issuers
Kerri Lemoie: 10. Open Badges VC: evidence
Kerri Lemoie: 11. Open Badges VC: Open Badges as a Skill
Assertion & Achieved Result
Kerri Lemoie: 12. Open Badges VC: credentialStatus
Kerri Lemoie: 13. Open Badges VC: proof
Received on Monday, 24 May 2021 22:57:25 UTC