[MINUTES] VCs for Education 2025-05-19 from meetings@w3c-ccg.org on 2025-05-19 (public-credentials@w3.org from May 2025)

From: <meetings@w3c-ccg.org>
Date: Mon, 19 May 2025 18:17:27 -0400
To: public-credentials@w3.org
Message-ID: <CA+ChqYdN6ko5Cf1SmaEnW3-hse-Z2R9HwwnFC88XgWYJ21Pn7A@mail.gmail.com>

📝 Meeting Summary: Quality Link Project Presentation

*Date:* May 19, 2025

*Presenter:* Colin Tück (Knowledge Innovation Center)

*Topic:* Quality Link - EU-funded project to create an open, interoperable
ecosystem for quality data on credentials.
Summary of Topics & Key Points:

*Project Introduction:* Quality Link aims to address the fragmented
landscape of credential data standards by building upon existing European
standards, rather than creating new ones. The project focuses on
micro-credentials and traditional degrees. It's a collaboration of five
organizations and runs until May 2026.
-

*Project Goals & Vision:* To create an architecture for publishing and
aggregating course information, trusted quality assurance data, and other
quality indicators using agreed-upon standards. This will benefit learners,
recognition bodies, and educational institutions, and support initiatives
like the European Universities Initiative. Alignment with Credential
Engine's goals was noted.
-

*Technical Architecture:* The project emphasizes using existing European
standards (e.g., Open Education API, European Learning Model (ELM),
Qualifications Data Set Register). Quality Link's contributions include a
management and discovery layer for standards lacking it, a data exchange
component for ELM, and ELM extensions for uncovered properties. All outputs
will be open source.
-

*Analysis of Existing Standards:* A review of various standards
highlighted differences in scope, terminology (e.g., "learning opportunity
specification" vs. "learning opportunity instance"), vertical structure
(number of layers in data models), treatment of learning outcomes, and
controlled vocabularies. ELM was chosen as the reference data model.
-

*Trust Model for Quality Indicators:* A three-tiered model was
presented: higher education institutions (data on their own credentials);
globally trusted sources (e.g., DECA, QDR); and other sources (e.g.,
ranking providers). Existing authoritative datasets will identify trusted
data sources. MOOC data integration was discussed, but currently excluded
due to lack of a comparable trust source.
-

*Next Steps & Engagement:* Draft technical specifications are available
for feedback. The project is building converters and a central aggregator,
with a pilot platform planned later in 2025 for testing. Several upcoming
events (Violent Tech Learning Impact Conference, UNICE conference, EDEN
conference, Quality Link webinars) were announced to promote the project
and solicit further engagement.

Suggested Next Steps:

- Xavi Aracil will consider mapping Ed API and ELM.
- Colin Tück will inform Ildiko Mazar about upcoming Quality Link
webinars.

Text:
https://meet.w3c-ccg.org/archives/w3c-ccg-vcs-for-education-2025-05-19.md

Video:
https://meet.w3c-ccg.org/archives/w3c-ccg-vcs-for-education-2025-05-19.mp4
📝 Notes

May 19, 2025
Meeting May 19, 2025 at 10:55 EDT

Meeting records Transcript <?tab=t.bluoe32v9cwy> Recording
<https://drive.google.com/file/d/1LlEs2Z9YDOpXXkXmvl53WHs4hrqywB1X/view?usp=drive_web>
Summary

Colin Tück from the knowledge innovation center presented Quality Link, an
EU-funded project aiming to create an open and interoperable ecosystem for
quality data on credentials by building upon existing European data
standards. The project's goal is to develop an architecture for publishing
and aggregating course and quality information, addressing the fragmented
landscape of data standards. Participants learned about Quality Link's
proposed contributions to existing standards, their analysis of various
models, and their trust model for quality indicators.
Details

*Project Introduction* Colin Tück from the knowledge innovation center
introduced Quality Link, an EU-funded project aiming to establish an open,
interoperable ecosystem for quality data on credentials, including
micro-credentials and traditional degrees. The project seeks to address the
lack of openly available and structured data on credentials due to the
diverse landscape created by micro-credentials (00:10:04 <#00:10:04>) (
00:12:45 <#00:12:45>). Colin shared that the project is a collaboration
of five organizations and will run until next May (00:13:56 <#00:13:56>).
-

*Project Vision and Goals* Colin Tück outlined the project's vision to
create an architecture where basic course information, trusted quality
assurance data, and other quality-related information are published using
agreed standards (00:13:56 <#00:13:56>). This would allow a central data
aggregator to consolidate and republish this data as open data, benefiting
learners, recognition bodies, and educational institutions. The project
also aims to support initiatives like the European Universities Initiative
by facilitating the creation of joint course catalogs (00:15:19
<#00:15:19>). Colin noted the alignment of their goals with the
Credential Engine (00:17:00 <#00:17:00>).
-

*Technical Architecture Approach* Colin Tück emphasized the project's
commitment to building upon existing European data standards for course and
credential information rather than creating new ones (00:17:00
<#00:17:00>). The goal is to identify and address missing links to
enhance interoperability, with a focus on flexibility and ease of adoption
for higher education institutions. All project outputs, including the pilot
aggregator, will be open source (00:18:19 <#00:18:19>).
-

*Landscape of Data Standards* Colin Tück presented an overview of the
current landscape of European data standards for basic course catalog
information, categorizing them into data models, data transport, and source
discovery/management layers (00:18:19 <#00:18:19>). They observed a
"patchwork" of standards with no single standard covering all three layers
comprehensively. Examples included the Open Education API, the European
Learning Model (ELM), and the Qualifications Data Set Register (00:19:42
<#00:19:42>).
-

*Quality Link's Proposed Contributions* Colin Tück explained that
Quality Link aims to contribute by proposing a management and discovery
layer for standards lacking it, a data exchange component for ELM, and
extensions to ELM for currently uncovered properties (00:20:54
<#00:20:54>). Draft specifications for these additions have been
released (00:22:03 <#00:22:03>). ELM has been chosen as the reference
data model for the project (00:32:02 <#00:32:02>).
-

*Analysis of Existing Data Standards* Colin Tück discussed their
analysis of various data standards, noting that while all cover programs
and courses, the scope of additional information varies (00:24:56
<#00:24:56>). Different standards also use varying terminology for
similar concepts, such as "learning opportunity specification" and
"learning opportunity instance" (00:26:29 <#00:26:29>). Notably, OKAPI
and DECA do not address instances (00:27:58 <#00:27:58>).
-

*Vertical Structure and Learning Outcomes* Colin Tück highlighted
differences in the vertical structure of data models, with some APIs having
a fixed number of layers while ELM allows for more flexibility (00:30:45
<#00:30:45>). Regarding learning outcomes, ELM allows for links to
external classifications, unlike some APIs that treat them as simple text
lists (00:32:02 <#00:32:02>).
-

*Qualifications Frameworks and Controlled Vocabularies* Colin Tück
pointed out that different standards refer to various qualifications
frameworks, though these differences are generally manageable (00:33:20
<#00:33:20>). A more complex issue is the variation in controlled
vocabularies for similar concepts, such as "learning schedule" (ELM) and
"mode of study" (OAPI), which are also modeled at different levels
(instance vs. specification) (00:34:59 <#00:34:59>). Efforts to
harmonize these vocabularies are being considered within the European High
Education Interoperability Framework (00:36:15 <#00:36:15>).
-

*Trust Model for Quality Indicators* In response to a question from Nate
Otto, Colin Tück detailed the project's trust model for quality indicators,
which includes three classes of data sources: higher education institutions
(providing data on their own credentials), globally trusted sources (e.g.,
DECA, QDR, able to provide any information), and other sources like ranking
providers (able to provide specific types of information) (00:47:24
<#00:47:24>). The system will leverage existing authoritative datasets
of recognized institutions to identify trusted data sources (00:48:50
<#00:48:50>).
-

*Integration of MOOC Data* In response to Ildiko Mazar's question, Colin
Tück clarified that the project is currently focused on officially
accredited universities and quality-assured providers listed in resources
like DECA (00:50:13 <#00:50:13>). While the technical framework could
accommodate other providers of massive open online courses (MOOCs), the
project lacks a comparable trust source for them and has therefore set this
aside for the time being (00:51:52 <#00:51:52>).
-

*Upcoming Events and Opportunities* Colin Tück announced several
upcoming opportunities to learn more about Quality Link and meet the team,
including the Violent Tech Learning Impact Conference in September (with a
co-located Quality Link meeting), a session at the UNICE conference in
Belfast, and dedicated Quality Link webinars on the draft specifications (
00:53:12 <#00:53:12>). Colin also mentioned their participation in the
EDEN annual conference in Bologna (00:54:57 <#00:54:57>).
-

*Call for Feedback and Future Engagement* Colin Tück encouraged
attendees to review the draft technical specifications and provide
feedback. They are currently working on building converters and a central
aggregator, with a pilot platform planned for later this year, offering an
opportunity for higher education institutions to test and expose their
course catalogs (00:37:40 <#00:37:40>). Ildiko Mazar thanked Colin Tück
for the presentation and invited them to share updates on the project's
progress in the future (00:56:03 <#00:56:03>).

Suggested next steps

- [ ] Xavi Aracil will consider doing a similar mapping exercise between
Ed API and ELM.
- [ ] Colin Tück will inform Ildiko Mazar about the announcement of
quality link webinars on the draft specifications so she can share the
information with the VC Edu community.

*You should review Gemini's notes to make sure they're accurate. Get tips
and learn how Gemini takes notes
<https://support.google.com/meet/answer/14754931>*

*Please provide feedback about using Gemini to take notes in a short
survey.
<https://google.qualtrics.com/jfe/form/SV_9vK3UZEaIQKKE7A?confid=Tx3P_hx81-pW8P4XcMLXDxIUOA8MCwMyBwiKAiAAGAMI>*
📖 Transcript

May 19, 2025
Meeting May 19, 2025 at 10:55 EDT - Transcript 00:00:00

*Ildiko Mazar:* Hello.
*Phillip Long:* Hello.
*Ildiko Mazar:* Hello. Good uh morning slash often. Not just good morning.
*Phillip Long:* I'm just returning from Spain and so I'm jetlagged.
*Ildiko Mazar:* Oh, I live there. So,
*Phillip Long:* Whereabouts?
*Ildiko Mazar:* you were in I I live near Jirona
*Phillip Long:* Uh
*Ildiko Mazar:* close to Barcelona.
*Phillip Long:* well
*Ildiko Mazar:* Did you go to Madrid?
*Phillip Long:* uh we were in Madrid but most of the time we were in the
southwest southeast
*Ildiko Mazar:* Okay. I wish
*Xavi Aracil:* What
*Ildiko Mazar:* I knew.
*Xavi Aracil:* should I do?
*Ildiko Mazar:* Hello.
*Colin Tück:* Hello. Hi. Hi.
*Ildiko Mazar:* Good to see. I thought you were on a train.
*Colin Tück:* Uh, no. Luckily, I managed to get off a bus uh just in time.
Uh
*Ildiko Mazar:* Perfect.
*Colin Tück:* um yeah. Uh I was uh hoping that it would work uh because
joining from a bus would have been a bit more fun. Uh
*Ildiko Mazar:* Yeah. Yeah. We we love adventures and uh it's it's nice to
prove that that technology can support when it can.

00:05:23

*Ildiko Mazar:* Okay. Um I see you perfectly fine and uh we have sound so I
don't suppose we need to uh have a technical check uh for those who are
already in the call. Uh Colin Tuk is going to be our guest speaker today
but uh let's wait a little longer and see some of our other regulars to
show up. We are still adjusting to the new environment. I might check
*Colin Tück:* Mhm.
*Ildiko Mazar:* into GTI to make sure that nobody's there, but uh
*Colin Tück:* All right.
*Ildiko Mazar:* we already did quite well last week, so hopefully people
will know.
*Colin Tück:* How come you changed from Jity to Google Meet? Was it a
policy decision or was it that Jity
*Ildiko Mazar:* Oh.
*Colin Tück:* didn't work well anymore?
*Ildiko Mazar:* Oh, now that you said that, no comment. No, it uh it it
kind of had a a partial liking of uh people on the opposite side of the
pond and the European colleagues were not transcribed. Uh oh, it was a
little difficult to have the uh full minutes of

00:06:39

*Colin Tück:* H.
*Ildiko Mazar:* meeting. Hello Dimitri. Good morning.
*Dmitri Zagidulin:* Hello
*Colin Tück:* So
*Ildiko Mazar:* We'll
*Colin Tück:* it
*Ildiko Mazar:* wait
*Colin Tück:* should
*Ildiko Mazar:* another
*Colin Tück:* negate.
*Ildiko Mazar:* minute or two, but I'm checking on GT if anybody might be
there by accident.
*Colin Tück:* back in a second.
*Ildiko Mazar:* Okay, very happy to report that there's nobody in Jity. So,
Colleen, you close the window and we're ready to roll.
*Colin Tück:* I'm just still making coffee as we speak, but we can uh don't
wait for
*Ildiko Mazar:* Okay.
*Colin Tück:* me.
*Ildiko Mazar:* Coffee.
*Colin Tück:* I'm ready.
*Ildiko Mazar:* Coffee is important.
*Colin Tück:* Um it's done.
*Ildiko Mazar:* Okay. Well, uh good morning and good afternoon to the
Europeans and welcome to today's uh PC Edgute Task Force call. The date of
today is the 19th of May, Monday, as per usual. So, uh just to start with,
uh some regular uh notes, call notes and IP notes. Anyone can participate
in these calls as you know. However, all substantive contributions uh to
these work items must uh be by made by members of the CCG community with
full IPR agreement signed.

00:08:34

*Ildiko Mazar:* Uh this can be found in uh the invitation and on our uh
GitHub uh site. Uh you have to have a WC account and uh you have to read
the uh contribution license agreement. Now we use these uh Google meet uh
links to conduct our calls. Uh these are all recorded and transcribed and
you can contribute to the conversation by typing in the chat or uh raising
your hand and uh joining the queue and
*Colin Tück:* Okay.
*Ildiko Mazar:* uh you will be given the floor if you have uh questions or
comments. Uh please keep your contributions brief so everybody can speak
and uh the machine generated minutes and the audio recording of every um
meeting will be archived on GitHub. With this said uh time to introduce and
reintroduce ourselves. Do we have anybody in the call who is not a regular?
I see one number that I cannot recognize, but the majority of our today's
small group I already am familiar with. And if nobody wants to take the
floor um time for announcements and reminders, uh is there anything um in
your necks of the woods or uh around uh your domain expertise or your field
of interest that you would like to share with the community?

00:10:04 {#00:10:04}

*Ildiko Mazar:* please speak up or type in the chat and if nobody wants to
speak then uh it's time to uh touch on our main topic of the day and hereby
I would like to warmly welcome our guest speaker Colin Tuk from the
knowledge innovation center who is going to introduce us uh to an EU funded
project called quality link There was already a small teaser in the meeting
invite. Uh but Colleen, you might want to uh introduce yourself and the
project at more detail. The floor is yours.
*Colin Tück:* Um yes, thank you uh thanks uh Leo and uh colleagues for the
kind invitation to join you today. Um so uh um my my name is Colin. I work
uh for the knowledge innovation center which is a small uh social
enterprise consultancy firm that specializes in innovation and
digitalization and education amongst others. I mean we have a um let's say
we quite strong regional focus on uh on on Europe and so has the the
quality link project that I'm going to introduce uh um a few words about.
Um, let me just uh get this few slides shared.

00:11:29

*Colin Tück:* Um, where's the button? Here's the button. Um, and yes. Uh,
okay. Now, just find the window that I need to share with you. Um, that's
the one. Okay, there we go. Um, I guess it's uh Wait, no, that's silly. Are
you now seeing this message that I'm sharing my screen? Oh, that's funny. Um
*Ildiko Mazar:* You never
*Colin Tück:* Um,
*Ildiko Mazar:* seen one before?
*Colin Tück:* maybe we should have maybe we should have done the check
before. I was um
*Ildiko Mazar:* I'm sure you'll
*Colin Tück:* uh
*Ildiko Mazar:* manage.
*Colin Tück:* let's try once again. Uh otherwise um h but that is strange.
I've shared uh slides in Google Meets a few times before but um it's
something has to uh something has changed it seems. Um okay how do we do
this now? Um
*Ildiko Mazar:* If you want to send me
*Colin Tück:* yeah.
*Ildiko Mazar:* something, I can uh try
*Colin Tück:* Um
*Ildiko Mazar:* to share my own my screen.

00:12:45 {#00:12:45}

*Colin Tück:* maybe that would be the easier one. Okay, let me try once
more. Um if this
*Ildiko Mazar:* Yes.
*Colin Tück:* ah yes now okay maybe I just clicked in the wrong place.
Okay. Um that's what I wanted to do before. Okay. So uh quality link is a
yes already said and you funded project um that uh generally tries to
create a um yeah open interoperable ecosystem for quality data on
credentials. I mean it does say in the title micro credentials um but it's
uh that's um let's say also a bit linked to the policy priorities that are
currently in place in Europe. In general, we are not only looking at micro
credentials but also at uh traditional degrees or any other kind of
education credentials. Um the uh the the reason why we're doing this is
that uh of course micro credentials have made the landscape a lot more
diverse and we saw that basic uh data on credentials is quite often not
available in open or structured formats. Just for the context, this is an
Arasmus plus funded project.

00:13:56 {#00:13:56}

*Colin Tück:* It's still running until next May. And you see here the
partners who are the partners so that's five organizations who are doing
this together um two universities uh uh the European university foundation
which is very active in arms without paper amongst others uh and KSC as a
European uh think tanks and technology organizations so to say and then we
have also Eden digital learning Europe which is another European
organization focused on digital uh learning um I'm running through this
first part a little bit faster so that we have some time to dive into what
we have found about um looking at different data standards on uh um basic
information on courses uh so far um but uh just to show what I mean what is
the vision we are working to how uh um we we want to create an architecture
where uh we have on the one hand basic information on on courses and
learning opportunities that are being offered. We have uh trusted data for
example from quality assurance bodies and we also have other maybe other
quality related information about the same course for example a rating or a
ranking provider who publishes some data um that is relevant when someone
wants to judge the quality of a microcredential learning opportunity.

00:15:19 {#00:15:19}

*Colin Tück:* The idea is that all these data providers can um use a set of
uh agreed standards to to publish their data in interoperable format and
that a central European or maybe in future worldwide data aggregator um
could uh could easily um aggregate data from various sources and republish
it as a open uh open data um all together. So you have all the um data
available to everyone in one single nice interoperable data set. Um of
course we hope that all this would uh um would benefit various stakeholders
and various use cases. So for example to guide potential learners but also
to inform recognition bodies in the labor market uh when assessing um uh
credentials or to um uh allow universities to benchmark with one another.
Um we also um see that there's a use case in general for of course
institutions having their learning opportunities visible Europewide and
there's maybe a specific European um policy initiative that uh plays into
this here is that um there's the European universities initiative that some
of you might have heard of and which encourages universities to form
alliances and work together towards a joint virtual campus and of course
creating A joint course catalog is part of most allianc's work program and
that of course uh is also uh something we want to achieve that that the
quality link architecture that we propose is something that alliances could
either deploy themselves or simply use to to manage their

00:17:00 {#00:17:00}

*Colin Tück:* joint course catalogs as well. Um the uh in uh as um I also
saw that uh colleagues from uh um credential engine have just joined the
call and I think I think what what we are trying to do of course is that
our goals and what I think what you're trying to do with the credential
registry are of course also quite similar and quite well aligned in general
I would I would say um um so uh what we tried when designing ing um um
proposal for technical architecture is uh maybe first and foremost we
wanted to make sure that we are not reinventing uh the wheel because there
are already quite a few um wheels around the European uh scene uh about
data standards and uh other thing related things for information on courses
and credentials. Uh so we we really wanted to make sure that we are
proposing some missing links where they um to for something to fill missing
links but not uh um to really build on on what is there and not uh um
propose yet another standard for something where we already have maybe
three or four different ones um which is something we'll then get to in a
minute.

00:18:19 {#00:18:19}

*Colin Tück:* Um, of course we want uh all this to be um uh as flexible as
possible and as uh with a as low bar for adoption as possible, especially
for high education institutions. So that we don't want to create something
that makes it super complicated for universities to bring their information
in. Um and of course uh it should be open also to future uh future
extensions. Um and um maybe something that for some would go without saying
but we want to underline it that what everything that we will produce also
the pilot aggregator of course will become open source software in the end.
Um so uh this is a bit a look at the landscape of what is already there in
um um in uh terms of European or Europewide used standards about basic
course catalog information and uh the exchange. Um we try to divide it here
a bit in three layers. Uh so that on the one hand we have a layer of data
models then of course we have a layer of transporting that data uh around
and then of course we have uh to be able to do all this we also have a
layer of uh discovering and managing sources from where to aggregate data
or from where to transport data for example into a central data set.

00:19:42 {#00:19:42}

*Colin Tück:* Um what we found is that uh we have a a nice let's say a nice
patchwork of various um data standards and also um API standards across
Europe that cover course catalog data in some way or another. Um but
there's not really anything that covers all three layers together and that
some some things cover some layers and some things cover other layers and
so on. So for example, we have uh the open education API or edio API and
the two examples which have a data model and the rest API transfer layer in
it but they don't have a uh built-in layer how to uh discover for example
uh an open API endpoint open education API endpoint. Uh then we have the
European learning model um which is a great ontology for uh all things
related to learning including learning opportunities. Um uh but of course
that is a pure data model anyway and doesn't have a transport layer built
in. There's a transport layer for the uh qualifications data set register
which is um something underpinning the European digital credentials
infrastructure.

00:20:54 {#00:20:54}

*Colin Tück:* But here that's a transport layer that is not accessible
let's say for individual high education institutions themselves but only
through the national authorities and so on and so forth. Um okay uh someone
needs to stop me. um otherwise you could spend half an hour on this uh
alone. Um so what we are trying to to build in quality link is basically
what is here illustrated as um red boxes. So we're trying to um propose um
generally uh usable management and discovery layer for uh various standards
that don't have that built in so that every known high education
institution could easily expose data in that format. Um and we also uh want
to propose a data exchange component for ELM data but that is not built in
the the standard itself. And then um we would uh last but not least um
propose some extensions to ELM for properties that are not yet uh covered
in the in the European learning model. Um and uh yeah so in the end anyway
there is also a link which I'll post in the chat as well later on.

00:22:03 {#00:22:03}

*Colin Tück:* um where uh you can find all the draft specifications for
these uh additional boxes already because we just released them a couple of
weeks ago. Um now what is um the let's say the main focus as we discussed
beforehand uh that I wanted to go into a bit more detail is what have we
learned from looking at the landscape of various data standards that exist.
Um I'll try to introduce them very briefly especially for the non-European
colleagues because I mean for maybe for the European colleagues this might
be a little bit boring because uh you've all heard of them but I think
others uh fortunately or unfortunately haven't. Um this is not necessarily
a fully complete overview of um data standards that are around in Europe
that cover uh basic course information. Um um but it's the ones that we
found to be at at least somehow uh let's say not necessarily Euro wide as
in used in all countries but at least um standards that are used at least
in several countries. I mean there are some national standards that are
used only in one single country possibly which um let's say we haven't
focused on.

00:23:17

*Colin Tück:* Uh so you have the European learning model already mentioned
it's um a data standard maintained by the by the European Union. So it's
the maybe so the most official of all of them. Um and it's used in the
digital credentials um European digital credentials uh infrastructure
amongst others. Uh and um it is um yeah RDF ontology and set of application
profiles. Uh then we have a data standard called Elmo which is used for
example in Arasmus without paper and the MAX network to exchange um student
and um course data. Uh then we have the open education API which is um
maintained by the Dutch national uh educational and research network called
surf which was originally a Dutch development but it's also used beyond the
Netherlands by now by some European university alliances and that is
basically um a standardized REST API that high education institutions can
expose to provide information on their uh courses and other things. Um uh
there's the open course catalog API which is or KPI which is um uh
maintained by the same organization that's currently managed the rustness
without paper network um that's used by some high education institution and
again it's a it's a rest API that allows institutions to expose their
course cataloges there's the um edu API by one tech which is used by
various institutions and um uh products in the future I guess uh um
although I

00:24:56 {#00:24:56}

*Colin Tück:* don't want to say something wrong here Jav but um um again
it's also a rest API uh standard that um specifies how cross catalog data
could be exposed by um institutions and then there's the database of
external quality assurance results which has its own data model which is
really focused on quality assurance um but it also includes some data on um
uh programs or courses that have been quality assured. Um now um what uh
what I'll try to do is highlight a few points of interesting especially
differences that we have observed in the work so far. Um uh I mean first of
all um they all cover programs study programs and courses um um in some way
or another um but almost all of them also cover something else um but
that's quite different what they cover so I mean some some of course are
very broad like the ELM is covering um uh loads of aspects related to
learning and learning opportunities. Um but then for example uh there's
some details that the open education API or the Edio API cover like for
example uh information on people being enrolled in a certain learning
opportunity which is in uh which is not the primary focus of some other
models or even details about rooms and buildings that are not uh at least
not uh covered to such a detailed extent in in some other data models.

00:26:29 {#00:26:29}

*Colin Tück:* Um so um I think this mainly illustrates that um it's uh
maybe on first side it might seem a bit absurd that there are so many
different European standards that cover course information in some way but
this explains a little bit uh we found at least that there's a a reason why
that is because um uh everyone is like a little bit coming from a different
perspective uh to look at course information and that might then explain
some of the differences as well. um there are um I think that different
covered I mean it's a it's a I really think it's an interesting observation
but it's not necessarily a problem as such I mean it's just something we
have to bear in mind that um uh course information and some standards is
linked to to other um uh things uh um depending on what else they cover uh
there's qu little bit of difference in terminology that different standard
uses use um which I think is also not really a a big problem. I mean takes
some time to get used to it but it's not it's not an insurmountable
obstacle in in general terms what we now use um to speak about learning
opportunity specifications which is a permanent thing and a stable set of
learning outcomes like a study program that is offered for years with a
stable set of learning outcomes so on and then where what we now use as a
term is learning opportunity instance for that program

00:27:58 {#00:27:58}

*Colin Tück:* being offered again every year or every semester or maybe um
it might be two instances per semester because maybe it's one university
that offers the program in two different locations and considers that at
different instances. Um yeah, the all standards use slightly different
words for that. Some call it spec learning achievement specification or
learning opportunity specification. Uh some just call it program cause
component. Um then they call the instance offering or instance uh so um
which is u um uh yeah um takes a bit uh time to sort the jungle but that in
itself shouldn't be a major problem. What is I think noteworthy as well in
this regard is that there are two standards uh okapi and also the deca
database that do not actually address instances at all. I mean all other
standards look at both the specification and the instance but these two
standards only look at programs or courses in general. Um uh now what we
did when diving into this in a bit more detail, we used the uh DEM data um
ecosystem data scheme echo system mapping tool um that uh maybe some of you
are familiar with and it's been developed by the T3 innovation network and
I think also used extensively by uh colleagues from credential engine um
which uh has uh yes I think some benefits for exploring the detailed um um
the yeah detailed differences

00:29:33

*Colin Tück:* of this data standard. Uh I mean first of all it is a nice
and uh easy to navigate user interface. So it's um it's yeah it's easy to
to browse around and filter your properties and it also scales relatively
well if you want to compare many standards which would make your Excel
sheet a bit huge. Um but it's maybe important to note then that uh um that
of course there are some some limitations or some things that uh um that
will at least will not solve for you if you want to engage in a similar
exercise. I mean um what it doesn't do is um directly um map or showcase
differences in the abstract classes or concepts. So if there are some
different understandings between what is an instance or um or what is a
specification I mean that's not so easily mapped. Um and also we found that
uh having different types of classes that partly inherit properties from
each other under that map to the same abstract class that that's not
necessarily easy to to represent. But most important of course it's a nice
tool for looking at the differences but there's itself won't build your
data converter.

00:30:45 {#00:30:45}

*Colin Tück:* So there's still a lot of work to be done um afterwards. Um
but it did help us uh explore some say some of the specific differences in
the um uh how things are modeled in this uh different um data models. Uh
one uh one issue that uh um will be an issue specifically when building
converses is actually the I mean I call this now vertical structure um
where where the open education API for example but also I think several
others like the edge API um are built on mainly a three layer structure
where you have some kind of a concept of a study program that is made up of
courses and then maybe the course might be made up of components um of
course that might um that's in a microcredential there there might not be
so many layers there might be just one but in a way there there many APIs
are based on the concept that there would be at most three layers um what
the European learning model on the other hand allows is that uh there um
there's an potentially an arbitrary number of layers uh so a learning
achievement specification may include many uh another learning achievement
specification which in turn might include another one.

00:32:02 {#00:32:02}

*Colin Tück:* So there could be five or six layers. I'm not sure that
anyone will ever um do that in practice, but uh they could theoretically.
Um so when uh we look at um converting data back and forth, of course, this
wouldn't really be a problem when converting from the right to the left
because um uh but it it might be a problem if you would want to convert the
other direction. Um uh this is fortunately not really a problem for um uh
for our um um uh for our use case or for our project because we actually
agreed that ELM will be our reference data model. So what we will be doing
is convert everything that is not ELM to ELM but will not convert from ELM
to uh somewhere else necessarily. Um but other people of course might need
to or want to do that at some point. Um we we also see that there are some
conceptual differences between the uh data models that are not necessarily
too difficult to overcome. I mean when you look at learning outcomes maybe
this is one of the the biggest area of differences where uh for example in
ELM you can uh summarize your learning outcomes as a text.

00:33:20 {#00:33:20}

*Colin Tück:* You can refer to the ASCO classification of skills and
occupations and you can also uh refer relate your learning intended
learning outcomes for cost to other frameworks or classifications. Um, and
some APIs like O API and OK copy it, uh, there's no built-in way of
relating learning outcomes to an external classification, but it's simply
like a list of text. And, um, uh, then, for example, in the ED API, there's
not the native spec, there's not um not necessarily a place to specify
intended learning outcomes. Um so so there are some some differences we see
similar differences at the qualifications framework and level of learning
opportunities where of course the ELM referred to the European
qualifications framework as it's an EU tool. Um uh it's not surprising that
for example the OPI being originally a Dutch standard refers to the Dutch
national qualifications framework which is part of the European framework.
uh and then um uh other standards refer for example edi has uh the general
concept of graduate undergraduate and doctoral levels but also allows for
extensions and uh then last but not least refers to the boner
qualifications framework um I mean these are these are more in the
qualifications framework sense I would say these are differences but they
are a bit more supple and they're not really difficult to bridge I mean for
example for our API defines eight levels in the same way as the European
qualifications framework.

00:34:59 {#00:34:59}

*Colin Tück:* So, so that is um very manageable. Um the last point of what
uh we've observed is that of course um different uh data models also have
their own controlled vocabularies and um there is um maybe uh um the
situation gets a little bit more tricky and and some of them I mean I
illustrated two of them here for example uh uh what's in ELM is called
learning schedule or in the O API is called mode of study it's um speaking
more or less about the same thing but it's um it's modeled slightly
differently and the control vocabularies and maybe not uh necessarily um
easy m to match one to one. Uh what makes it maybe a bit more tricky for
this one is even that in elm if I'm not mistaken learning schedule is a
property of the learning opportunity instance whereas in the O API it's a
property of the learning opportunity specification. Um so uh if you would
have the same program that's offered in part-time and then full-time and
that would be just two instances of the same program in Elm, you would need
to make this two completely different programs in our API.

00:36:15 {#00:36:15}

*Colin Tück:* Um uh there's a uh similar situation when it comes to um uh
the mode of delivery or mode simply mode and ELM uh which at least in both
cases it's at a level of specification but uh again the two controlled
vocabularies are um looking at very similar uh topic uh with um slight
differences Um so uh this is something that I mean we observed in in the
project and we would uh somehow have to manage for at least for the sake of
building converters from O API to elm for example but it's also one of the
observations that we are trying to feed into current work on the um
European high education interoperability framework that is managed by the
EU which has uh for example a task force that is um specifically dealing
with data uh standards and APIs. And one of the things that might be
proposed there is for example uh some initiatives to harmonize controlled
vocabularies across standards for for these type of common um common
matters. Um so that yeah there's still some some work to be done. Um just
to wrap up with where we are standing with the quality link and what is
also next.

00:37:40 {#00:37:40}

*Colin Tück:* Um so as I mentioned before we have um a set of uh draft
technical specifications out for consultation since a couple of weeks. So
we would be really excited if you have a look at them and uh share your
feedback on on them with us. Um so there which cover all the um boxes
mentioned before. Um there's also a survey on on quality indicators that
might be of interest to you. And what we are working on next is actually
building um converters between different data standards under a central
aggregator um based on those uh draft specifications. And of course that
means uh at some point later this year this um converters and aggregators
will be uh will be functional and will be also um what the aggregate will
become visible in a pilot uh platform um and at least for for high
education institutions that means uh there's an opportunity to join and
testing this and also expose your own course catalog and the in the pilot
platform then um which uh again will be um done in a way that every um uh
at least every uh European high education institution could easily uh
expose their course data in ELM or O API or edio API format for example and
have that aggregated without uh having to um make some anyone else do
anything else than what they need to do.