<marty_reed> Good morning, Marty Reed with Randa solutions.
<kimhd> Kerri, can I call on you to reintroduce yourself.
<kerri_lemoie> My name is Kerri Lemoie, and I'm from open works group and I've been working in an open badges and EdTech for a little bit over a decade. And now I'm transitioning from a technologist to do more research. So I need to know everything that's going on all the time about the standards so I can like understand it and do the best research.
Topic: working session modeling verifiable credentials for education
<kimhd> One thing I wanted to start with was
<kimhd> As we progress further we're developing several artifacts. One of these is the VC EDU models spec. And I think, to the discussion last week, a lot of additional topics have gotten pulled into the draft spec that we can possibly move outside other work items of the task force and actually let me go ahead and include the direct link.
<kimhd> First, there's a high level topic of scope. Representing skills, some more formal education, like a diploma and then informal education like webinar attendance.
<kimhd> Within that there are several issues that would end up being maybe too low level. So for that work items -- possibly out of scope for that document. So some things might include the XML mapping work that we have to do to make verifiable credentials usable as XML and actually I can paste these in as up potentially out of scope.
<kimhd> And it's possible Anthony's PR will address some of this. So maybe I'll come back to this. So I would possibly propose that out of scope for this document will include things like details of the XML mapping, because that's going to be a PR against the VC maintenance charter group.
<kimhd> General solutions for things like image integrity. So this is something we talked about last week, which is people will be representing images in different ways, whether it's HTML, SVG, PDF, maybe links to an image that's hosted somewhere, but then also there's the concept of integrity of that image so general solutions for things like that might be out of scope. Oh, and then also PDF as a container I would recommend is out of scope. I think what I'm getting at here is that we might think of what we're focusing on in this initial document is more a descriptive versus some of these other things get into more prescriptive and then those are the things that will require their own efforts to iterate around like image integrity. How do we represent the general notion of that and we're not saying prescriptive.
<kimhd> For a broad set of of people using verifiable credentials, just in terms of if you want to maximize interoperability for wallet users, things like that. These are our recommended practices. So I'm just going to add those next. Let me turn to Anthony's PR
<kimhd> Anthony, can I call on you to talk to this.
<anthonycamilleri> So if you could just call it down a little bit more. So the green part can all be in there are sorry for getting this everyone so late.
<anthonycamilleri> So what I tried to do
<anthonycamilleri> To do was describe what at least from my point of view, are the basic concepts we need to include
<anthonycamilleri> I very purposely didn't name the concepts, so that we wouldn't,for example, have problems about what does credential engine call a credential versus W3C calls a credential, etc. If we agree on the concepts we can name them anything under the sun.
<anthonycamilleri> To my point of view, there are four concepts that we need in an educational space. The first concept describes an activity, somebody has done.
<anthonycamilleri> They performed Beethoven's six symphony, they attended a conference, they spent six years working as a lion tamer
<anthonycamilleri> Things they have done and it's really just things this person has done -- that's it.
<anthonycamilleri> Secondly, the second concept.
<anthonycamilleri> Would be a learning outcome or a skill and well you know how long people in education argue about difference between those two. And I really don't want to get into it.
<anthonycamilleri> But examples of this concept would be piano playing, advanced woodworking, or if you imagine something in a more learning outcome Format, work in accordance with established sustainability. But these are all the learning outcomes skills, etc.
<anthonycamilleri> Um, the third concept basically describes the outcome of a formal assessment process. And it's very much oriented toward formal education and examples of this bachelor's in sociology a Microsoft certification.
<anthonycamilleri> A certificate that says you're a member of the bar. These kinds of things,
<anthonycamilleri> And then the fourth one described a person's participation or membership in an educational activity. So an example might include that you are students of the University of Wine country, you're an apprentice lawyer with the California Bar Association, etc, etc.
<anthonycamilleri> Things to note about these concepts. First of all, all these concepts can be related to one another. So, for example, a bachelor's in sociology
<anthonycamilleri> can include the acquisition of the skill of sociological research, it can have been performed after
<anthonycamilleri> 500 hours of fieldwork practice while being a student of the University of Orlando. Ah, so let's say all of these concepts quoted as saying include relational fields to each other.
<anthonycamilleri> The second part of this is the approach to defining this and let's say this is kind of a mix. It's a bit of a hybrid of the examples that were provided below. So, what you find is that
<anthonycamilleri> You could talk about all of these concepts being able to be described in terms of a definition and instance. So a definition is a standardized description of that concept.
<anthonycamilleri> It's a link to a URI describing the concepts and a type field or framework URI to say where it's coming from.
<anthonycamilleri> So a badge class would be a typical example of a definition for example and then because I felt some models needs things you would also have an instance
<anthonycamilleri> Awesome and an instance would describe but data about the person's individual performance of that concept. So if you're talking about, I don't know, and Bachelor in Sociology. That's where you would put the person's grade or the dates the person that went to that course. And they all the stuff that is unique to the individual.
<anthonycamilleri> And is associated with that concept. If you scroll a little bit more down. I put a very, very basic example in JSON, which you should see there. So here I put
<anthonycamilleri> In line 19 you see schema hasConcept is like I just put hasConcept, so I don't get into the terminology and you see that the definition has an ID.
<anthonycamilleri> And a framework and there would probably be several frameworks accepted globally. And then those are the only two you have
<anthonycamilleri> We could have a whole descriptive, then probably want to include the name and description, but let's say
<anthonycamilleri> What goes below the concept id and framework that's a discussion for another meeting. And then below that. There's the instance of that definition.
<anthonycamilleri> I just invented a couple of fields here and send data performed or grades, but again, let's say what goes into the instance would depend on, let's say, which educational standards, you're using
<anthonycamilleri> So those are, let's say the first high level thoughts I had about this. Again, this is really meant to just be a framework or template for discussion and to get the ball rolling.
<kimhd> Fantastic. I really like this. There are a couple of things I wanted to... Okay, I don't see anyone's hand raised.
<kimhd> Wwe don't have that many people, so feel free to just interrupt me if I'm missing you, but one thing that I really like about this is this hasConcept definition/instance has. So one thing that I've been struggling with in modeling is the idea of
<kimhd> And, you know, I achieved some credential according to that framework, right. So this let's use this agency facilities management as an example. Then there are things specific to my instance of achieving or acquiring that
<kimhd> Because the way I've seen a lot of them model. It is the images are being put under credential subject which doesn't seem right. It's not a picture of me the subject. It's a picture of my instance of award of that credentials. So images in here would be like a static, you know, something that every recipient of that credential of agency facilities management would get. Images here could have like my name embedded in it or whatever those specific to me. And so this is the first example I've seen that kind of tease that apart, independent of what we call them right now. I like this a lot.
<kimhd> Let me see if we have a queue right now.
<juan_caballero> I just had a quick question. If
<juan_caballero> Maybe this is just following on the same lines, but I was just wondering is if there's any sense of like PII goes here or like no PII go through like what the instance. Could it be some record systems for for student credentials would have, like, I don't know, a student identifier in there or
<juan_caballero> Just wondering about that because Kim was saying the image that might contain PII or
<juan_caballero> something specific to the student would would happen in the instance section is that sort of by design.
<kimhd> That was just an example of how you could use instance in the API question. Is it big and it will be going on, we need to. So I'm thinking that for this job spec, a more complete survey of that may be out of scope but with that.
<kimhd> We definitely need to include in the privacy, security concerns what's happening there. But I do think we need you know basically like the concerns like you don't want to put it on a blockchain or things like that. But we do need to
<kimhd> I mean, that's a general problem with verifiable credentials. But I think that it would be useful for us to push forward on that specifically in the context of education. I'm curious, maybe I could ask
<kimhd> Well, actually, I'll get to Phil first and then I was curious to ask and Anthony if what's happening in EDCI, for example, Phil your up first.
<phil-t3> Thanks, Kim. I'm trying to get my head around the, the level of the conversation here because it sounds much to me at least, like what Anthony suggesting has many of the elements that are in, for example, the the data model for badges, an assessment pointer to the method by which a particular achievement or or capability was determined
<phil-t3> Or evidence of that determination in terms of an array of things that a person might have done that they can be pointed to, so I'm just trying to understand how what we're talking about here differs from the description of other data models that have been put forward for carrying a credential.
<kimhd> I can talk to that quickly and then I'll work through the queue.
<kimhd> Which is that I think what we're talking about here is the more generalized notions. So I think that, you know, there are other say other non Open Badges type schemas. This one specifically we're trying to avoid use of certain terms. We're just talking about the concept of it and then those can be mapped to these other frameworks, but there's a mapping work in T3
<kimhd> In you know this credential credential here means credential there, this field here means that field there, I think what we're talking about is slightly higher level than that we're talking about, you know, what are the the concepts we want to express, and what are the general categories we want for it. So like this idea of, the definition versus the instance and then I don't think it will necessarily be saying like things have to comply with this structure like this isn't prescriptive for Open Badges to say adopt the structure is useful for understanding how what they have in Open Badges map to this. That's my personal take on it
<kimhd> Add yourself to the queue if you have more. If you have additions carry your up
<kerri_lemoie> Actually I don't have much more than Phil because I was thinking the same thing that we're talking about someone like an individual's assertion.
<kerri_lemoie> And the corollary it Open Badges with like, you know, a baked badge, the badge image. What could have something that is personal and unique, whereas the badge image overall is something that you know all the recipients references. So, um, and also, I guess I should say that in the out of scope.
<kerri_lemoie> Issue that you had up earlier because I was a bit late. Sorry, I am I added something about evidence and documentation. So I don't know, maybe if that fits in here in terms of like providing evidence of some kind.
<kerri_lemoie> Of proof, visual, or audio or some kind of media reference or something like that. I was I'm probably too specific at the stage so
<kimhd> Oh no, that's interesting. So for evidence are you seeing that would be out of scope.
<kerri_lemoie> Well, I put it in the out of scope document ask that question to see if it is out of school. I wasn't sure if it come up yet, but it's a pretty critical aspect of matches that I think has been under utilized that could be more utilized
<kimhd> Yeah, that's a good point. Putting it in the issue is a good start. I'd like to continue iterating on that I
<kimhd> Think. Yeah. Actually, we'll come back to that. I was going to offer some opinions on that but I
<kimhd> Yeah, we'll come back to that. Let me see what else I might be missing data ecosystem schema mapper is that the term that I was looking for earlier?
<phil-t3> Basically yes, that's the link that you were talking about the T3 that use you pick a spine as a reference. And then you map directly to the elements of other representations
<kimhd> So this is in sort of the intro section where we're talking about what is the scope of this documen, and then we have this concept example.
<kimhd> Now, Anthony, can I ask you to comment on how you see the this informing the rest of the document so
<kimhd> My read again is this is talking about conceptually the generalized structure. And then this specific examples we give you know may not necessarily comply with this, but this is a good generalized structure for things like where we're bridging the frameworks, like something defined in credential engine would be a really good example of that.
<kimhd> That's Not trying to describe the instance; it's describing more what's in this definition part
<kimhd> So, yes. Any other comments on, you know, sort of context and purpose of these additions.
<anthonycamilleri> Maybe I will answer to as well here. So we're clearly not trying to build a empire edu standard here.
<anthonycamilleri> Let's say, for example, 4 concepts was an attempt to say are these the concepts that can map across all the various educational standards. If an open badge could fit into a verifiable credential, the EUROPASS would fit and the credential engine would fit and that would at least be like a small base of concepts that can be carried over between them, and that was also the reason I put in definition and instance, is because if you look at the European ones are in particular EMREX and Europass this instance is a big deal mainly because they were built for formal education and by the way, the instance terminologies also stolen from there, but I'm not wedded to it in any way.
<anthonycamilleri> But moving forward from here. The idea would be I think that first we would find the names for these concepts which we can agree on. And I'm very afraid of that process. Actually, I'm
<anthonycamilleri> But then, secondly, we would need to, I think, agree, what is the minimum amount of data each concept should contain, which I think is the trickier one -- a name, we can all agree on beyond the name, it starts getting more complex. And I think we need to discuss
<anthonycamilleri> In terms of what goes into the instance, I don't think we need to define anything at this stage that goes into an instance
<anthonycamilleri> I think it would be lets say up to the framework to define what goes in the instance. There are a couple of things you might say, we would consider standardizing you could consider putting the grades there, you could consider putting an evidence there.
<anthonycamilleri> But let's say for both of those are. I have a question mark in my own head, whether let's say something we should deal with here or if that should let say belong somewhere else in terms of standardizing and originating
<anthonycamilleri> So those are my thoughts are so far. And to be honest, further thoughts will come. Basically what people say this cool
<kimhd> Great Phil Barker, you're up.
<philbarker> I'm looking at this and I'm trying to get my head around whether it fits with the current things that are already in schema.org
<philbarker> And hasConcept I'm finding a bit confusing. And I'm wondering if there isn't some way of using the existing hasCredential. Measuring the sort of thing that's already in schema.org between a course and a courseinstance to have a credential and a credential instance and that perhaps skipping the definition thing because it, you know, it seems to me that the definition is what you already have for the range of hasCedential, which would be a credential described in JSON LD.
<philbarker> Which sounds a bit rambling. I just wanted to say that if this seems confusing to me, given the background of what's already in schema.org
<kimhd> I'll comment Anthony, but my impression is not that it's meant to replace that. I think this is just a kind of meta, meta model. I think right so hasConcept...
<kimhd> He intentionally doesn't want to put word specific words in there, right, because there's discussion about like hasAchieved versus hasPerformed vs hasCredential. And I think that we're, we're all aligned that hasCredential is it's good that it's there, we can reuse that
<kimhd> But the idea of what are we trying to describe in general before mapping to some specific things. So Anthony, can I call on you to talk more about that.
<anthonycamilleri> Though I think you covered it, the issue with hasCredential is, I don't know if hasCredential would apply equally to have four concepts I describe above
<anthonycamilleri> If so, then you might use hasCredential and make those types or those could be four different high level concepts. So this isn't something I am clear on myself, either at this point.
<philbarker> Okay, that's fine. That's fine. I was just getting a bit confused with the sort of the point where the way wording goes from this vague hasCredentials to the very specific example underneath that. But okay, I understand that the context. You've just explained
<kimhd> So one actually. Other thing related to this is that we probably need to get editors so, my comments on editors versus authors. So my understanding is for editors.
<kimhd> That's more, the people that are editors are also authors, but then they're also taking responsibility for doing PRS and accepting commits to the spec and then also just sort of they are more owning the general direction of the spec.
<kimhd> And then authors. I think that there's some heuristics that are used for that. But typically, I think.
<kimhd> You know, it's something that self identified; if this spec is something you want to contribute to you see yourself making
<kimhd> commits in the future, then we can add you as an author, and I think that there are few people that I've identified as, you know, sort of contributing to this already. And feel free to can can people just reach out to me directly. If you're interested in being an editor versus author, and what I'll do is get an initial set of names on the spec itself, because I think we do need to have some initial set of editors to approve PRs just so it's not rubber stamping this, but I think that my proposal would be that we accept this. And I do think possibly by the way I'm showing it on my screen like it's hard to keep the context like Phil was talking about, for example, like we don't want people to read this and think that we're proposing.
<kimhd> That people write credentials, like with this hasConcept term, we want to make clear that
<kimhd> The framing of what we're talking about here, but I think that's more editorial versus this content, which I think we now understand the context for, so
<kimhd> I would like to propose that we merge it with those clarifications, if needed, but then also we should have editors and authors (and reach out to me if you're interested in that.)
<kimhd> We have some discussion about K-12 assertions, Marty. Do you want to talk to that?
<marty_reed> Yeah, so we've been we've been working on kind of the CLR JSON model and I was just wondering, were capable of assertions like GPA or even a know base64 encoded version of the transcripts might fall in these in these concepts, because those assertions seem to overlap these concepts. And so I was just kind of wondering if there was a specific one that that you thought about or is that separate
<kimhd> It Anthony, you mentioned concept three. Do you have anything to add to that?
<anthonycamilleri> Yes. So certainly are anything that's the result of some kind of assessment, generally we consider as concepts free for the moment.
<anthonycamilleri> Now, if the transcript makes references to learning outcomes achieved as well. You might say concept three with references to concept 2. But basically concept three is meant to be any type of diploma, transcript, or any other typical paper formal education institution.
<marty_reed> So for me the language there is just a little confusing when you say formal assessment process because that I guess it could be considered a course is andassessment of skills, but just some way to clarify that language a little bit would be helpful.
<marty_reed> A formal education process or I don't know how what the right terminology is but, the term assessment from a k-12 perspective is very specific ersus you know, a course, per se.
<marty_reed> Are we getting too granular here?
<kimhd> No, no, that's good. There's two. So there's two paths to what I'm thinking right now. One is that so, so I think this framing of like what concepts we're trying to address is critical. I think then, within that, we're going to call out there's a few specific examples that we want to demonstrate.
<kimhd> We haven't yet locked down what those examples are. And so I think, you know, for example, if you want to propose this k-12 example, that might be a good one, especially if that's something that's happening in LER pilots.
<kimhd> So there's there's one thing that we might actually take a stab at mapping that out in the context of this document, but at the same time, I think for, further refinement of that language. I'd like to land this PR and then keep iterating on that. So, for example, we could either open issues or PRs and come back to it in and other discussions.
<kimhd> I don't know if there's an easy change to that. But I think that we have some people here who could help refine that language. So I'd recommend opening an issue in on the VC-Ed models repo to sort of clarify how k-12 terminology fits in.
<marty_reed> Makes sense. Thank you.
<kimhd> Joshua was that I don't know if that was a comment. I'm not sure struck means
<phil-t3> I think what I what I think he's trying to do is is I described in the chat is lay out a sort of a superstructure for this conversation, within which more detailed things need to fit or for which, if they don't, the superstructure needs to be questioned and the one thing that I didn't notice in here is that Non formal or informal learning activities.
<phil-t3> I'm not sure quite fit in the superstructure at this stage.
<phil-t3> Because they can be every bit as
<phil-t3> As evidence based as
<phil-t3> A structured thing they're just not presented in the context of the learning outcome intended there. They're often occurring in work and various other learning focus structured activity.
<joshua_marks> And then in addition, there's a bunch of information that occurs in the transcript that describes the context of the achievement, but not an achievement in and of itself.
<joshua_marks> You know, like, what term in which the achievement was achieved or the time period that the credential was awarded, etc. So there's many things that I don't think fall into these concept buckets that are important to something like a transcript.
<joshua_marks> And have been formalized in many transcript structures. And I think what Phil was saying also goes along with that, there's many things in job context that are contextual or summary that don't fall into these buckets.
<kimhd> Anthony. You're on the queue.
<anthonycamilleri> Okay. Ah, so what you're referring to Joshua.
<anthonycamilleri> I fully agree, and I understand.
<anthonycamilleri> I mean, I think this is something we could get in terms what Phil was saying I could, I asked you for a little more explanation in the sense that for me.
<anthonycamilleri> In terms of an example because for me this is what concept one was for, to describe informal learning where you say, I just did something. It wasn't part of an educational program or so on. Here's something I did. Here's my work experience. It's just something I did.
<anthonycamilleri> Or you could use conception to say I have this skill, I'm not telling you anything about this program, etc. So I'm not quite understanding what's missing is something is absolutely no issue, but I need a couple of examples to get where you're coming from.
<kimhd> Yes, Phil.
<phil-t3> Thanks, Anthony. I think the main crux of what Joshua was alluding to this is that the context. I think seems to matter. One might, for example, in the context of working on an accounting system in a job become facile with particular methods of calculation with that involve statistical calculations of one sort or another and learning it that way, as opposed to being in a course in accounting that specifically intended to build a structure for you to learn those sorts of things to me at least implies a different a different set of experiences that, on the one hand, both lead to the ability to do a particular thing.
<phil-t3> But the non-structured approach conveys that that was done because the person brought to it's something that they were able to leverage to achieve that outcome as opposed to having that constructed in an intentional way for them to be able to build on.
<phil-t3> Don't know if that helps you
<anthonycamilleri> I know that sounds completely on what I would suggest is, either you or I could do it, I would suggest that we open an issue for the GPA example and for this exam I will keep the other use case. So, and then, basically we can give it a shot and trying to break it down
<kimhd> Yeah, that sounds good. Thank you for clarifying that.
<kimhd> Okay I think PII might be too meaty to get into now.
<kimhd> We might have that as a separate discussion with maybe with a little more preparation. Anyone have anything else to add, we might be able to end a little bit early. If not,
<kimhd> Phil Barker?
<philbarker> Yeah. Could you scroll down a wee bit on the screen, please. So it's part of what I covered in my last ramble
<philbarker> I think what I'd like to see in this and what, you know, if you like, I can you know if this is accepted as a public rest, I can build on that is that the the schema hasConcept shouldn't go to something that's got a definition, it should go straight to the definition of the thing.
<philbarker> And the instance should be an instance of a thing. So if you have schema hasCredential.
<philbarker> Then what you have underneath that is the idea of the credential and type credential, which is already a typing schema and then
<philbarker> Consider adding to the credential class in schema.org a hasInstance property which goes to what's listed here as a separate property instance, because what I'm not seeing here you know types. I'm not seeing domains that match what's in schema.org at the moment.
<philbarker> I can imagine a hasCompleted that goes to a course and then you can use the existing course instance to point to the particular instance of the course.
<philbarker> So I don't think I want to try and explain that any further in words, but if this is accepted as a pull request. I can try editing it to show more clearly what I mean
<kimhd> One thing I'm wondering is, maybe having the schema prefix is sending our heads in the wrong direction, because I think I'm seeing this, not necessarily as something that someone would express directly. So Phil B I definitely hear what you're saying. Like in terms of how schema org works.
<kimhd> You would have some of this in the type; I think that this is just calling out a meta structure again and and i think one way we might do that is just expressing it more like in pseudo code, rather than in JSON; just making it clear, like this is this is not a thing that we're recommending you go off and do this is the, these are the high level elements.
<kimhd> Then, in terms of how it how it's actually implemented, I think, is what I'd really like to get that from you because I think how that maps to schema.org -- what's in there right now. That would be very useful.
<kimhd> But I could take a stab at that, if if Anthony is all right. Obviously I'll submitt as a PR. But Anthony What do you think about that, like if I sort of rework it into kind of more pseudo code kind of
<kimhd> Less weight then JSON-LD. And then maybe the interpretations people are taking with it.
<anthonycamilleri> No problem from my side, anything that makes it clear, and let's have some fun debating.
<kimhd> Any other comments before we end the call.
<kimhd> So I want to thank everyone. This was extremely useful for sort of start working through some of the issues we're going to continue having these working sessions and, you know, focusing on this and any related efforts. learner wallet discussions might come up more in the coming weeks as a separate issue, but it would be nice to get this thing closer to landed and finished off or at least at a stable point where we can continue iterating in the background.
<kimhd> So Phil reiterate next steps. Good idea.
<kimhd> We have a lot of them. First we need editors and authors and they don't have to be the final set but I'd appreciate if people could reach out to me directly. If you're interested in being an editor or author just email me.
<kimhd> And so we need, we need some names on the spec. The editors are the people who should review the PRs as they come in. And so this will be a good forcing function.
<kimhd> Let's see. Okay, so then we get editors, we merge this, and then we start reworking into some, something that captures this discussion. So, for example,
<kimhd> It seemed like the framing of it, you know, describing as JSON-LD versus something more pseudo code type of thing was maybe confusing people. So I'll take a stab at that. There were a few other things that came up
<kimhd> For example, Joshua Marks had a good term for it. Summary outcome or something, he was using GPA as an example.
<kimhd> Were in our concepts we didn't quite have a category for that yet. So I think that I would appreciate it if we could get filled long in Joshua marks to hopefully add some details there. So we'll probably need to do that in a in a tracking issues. So I can send that out.
<kimhd> Oh, thank you, Phil, you're writing them to me. So the the other one would be creating an issue around summary outcomes, perhaps
<joshua_marks> Yeah. Summary outcomes and outcomes context.
<joshua_marks> So the two buckets or the outcome context is gonna be tricky because it occurs at all levels of the hierarchy.
<kimhd> And then there was one other thing that I just remembering from way further back in the discussion is Phil Long mentioned around recommendations about use of DIDs.
<kimhd> There's several layers to it. But I do think that it will be useful for this document to have something around at least a descriptive definition of how DIDs can be used.
<kimhd> Because I think I'll be able to provide that
<kimhd> Any more thorough examination of which did are useful for which kind of context would be out of scope, but I think we could bring something like that in scope. So that's the last thing that we had privacy. PII will will definitely come up again. So I don't, I don't know if there's an issue we need to create right now. But we'll, we'll come back to that.
<kimhd> Great. Thank you so much, Phil. Saved me some time.