Accreditation discussion on Schema.org list (public-vocabs@w3)

[Greg Grossmeier]

See http://lists.w3.org/Archives/Public/public-vocabs/2012Oct/0006.html

The gist of the proposal:

> I think schema.org could do with an Accreditation class. This should
> be an Intangible object, and property of at least Organisation,
> Person, Product and Place.
>
> The properties this class would need as a minimum are:
>
> Accreditation name (text)
> Accrediting body (an Organisation)
> Level or Grade (text)

Seems to have received positive feedback thus far,

Greg

--
Greg Grossmeier
Education Technology & Policy Coordinator
twitter: @g_gerg / identi.ca: @greg / skype: greg.grossmeier

[Sean Connor]

Hi Greg, very nice!

I read through Nicholas' message...would courses of study (or individual courses) also qualify as "things" that carry accreditation, or is that already handled separately in LRMI/Schema.org nomenclature? For instance, individual courses/exams can be recognized by the National College Credit Recommendation Service (NCCRS) for transferable credit; that recognition is particular to a well-defined thing: the course/exam package.

Sean

[Steve Midgley]

This seems like very good news. My two cents would be that alignment with OBI/Badges could be helpful also (since there is some discussion of awards vs accredition vs authority, which they have to deal with also). Getting OBI into a pure schema.org structure for representing their data would be a big win for everyone, and this might be a start?

Steve

[Chris McAvoy]

On Fri, Oct 5, 2012 at 1:42 PM, Steve Midgley <steve....@mixrun.com> wrote:

This seems like very good news. My two cents would be that alignment with OBI/Badges could be helpful also (since there is some discussion of awards vs accredition vs authority, which they have to deal with also). Getting OBI into a pure schema.org structure for representing their data would be a big win for everyone, and this might be a start?

The OBI structure for a badge is described here. The structure of the assertion is straightforward,

{
  "recipient": "sha256$2ad891a61112bb953171416acc9cfe2484d59a45a3ed574a1ca93b47d07629fe",
  "salt": "hashbrowns",
  "evidence": "/badges/html5-basic/bimmy",
  "expires": "2013-06-01",
  "issued_on": "2011-06-01",
  "badge": {
    "version": "0.5.0",
    "name": "HTML5 Fundamental",
    "image": "/img/html5-basic.png",
    "description": "Knows the difference between a <section> and an <article>",
    "criteria": "/badges/html5-basic",
    "issuer": {
      "origin": "http://p2pu.org",
      "name": "P2PU",
      "org": "School of Webcraft",
      "contact": "ad...@p2pu.org"
   }
  }
}

There are fields that are specific to a badge, like the graphic, but otherwise it ties a learner (recipient) to a specific curriculum (badge:criteria) and an accrediting organization (badge:issuer).

I'm part of the OBI team, we're beginning to encourage badge issuers to use LRMI in their criteria pages, it makes sense to do a schema.org compatible badge description as well. My initial reaction is to use our badge assertion as a basis for the schema.org description of an award / accreditation, then use a super set to describe a specific badge (so things like badge image become optional to the schema.org description, but required for an OBI compliant award / badge).

Really glad to see the conversation get started...

Chris McAvoy (ch...@mozillafoundation.org)

 

Steve

On Thu, Oct 4, 2012 at 9:17 AM, Greg Grossmeier <gr...@creativecommons.org> wrote:

See http://lists.w3.org/Archives/Public/public-vocabs/2012Oct/0006.html

The gist of the proposal:

> I think schema.org could do with an Accreditation class. This should
> be an Intangible object, and property of at least Organisation,
> Person, Product and Place.
>
> The properties this class would need as a minimum are:
>
> Accreditation name (text)
> Accrediting body (an Organisation)
> Level or Grade (text)

Seems to have received positive feedback thus far,

Greg

--
Greg Grossmeier
Education Technology & Policy Coordinator
twitter: @g_gerg / identi.ca: @greg / skype: greg.grossmeier

--
@chmcavoy
http://lonelylion.com

[Carla Casilli]

Hey all,

Really happy to see this conversation progressing! A note about the term accreditation: Chris has pretty clearly outlined how we'd like to use this concept—as representation of the issuing organization. This usage may contrast with typical expectations about accreditation, which generally is associated with formal higher academic institutions being acknowledged by formal accrediting bodies as meeting specific standards. Since Open Badges seeks to address both formal and informal learning and we do not have badge accrediting bodies, we'd like folks to be aware of the nuance in that proposed term.

If you have extra time to spare, you can read some of our additional thoughts on validity, credibility, reliability here, Badge System Design: what we talk about when we talk about validity

Thanks!
Carla

[Ross Higashi]

This is indeed something that I've been wondering about with regard to OBI and LRMI. However, moreso than just making sure that badges carry the metadata to assert accreditation, I'm wondering how to handle the business of credibility in making that assertion.

Given that the selection of alignment claims at both the OBI and LRMI levels is determined by the tagger, it stands to reason that a content creator doing self-tagging could do a poor job of alignment, make overly strong claims about a weak activity, or even make fraudulent claims altogether.

Is it within the scope of either OBI or LRMI to include metadata about the strength or credibility of an alignment claim, for example by allowing the tagger to designate third-party organizations that might accredit or endorse the claim? Or does this task fall entirely on the shoulders of the recipient of the information, with these systems simply being a neutral reference so that everyone can agree on what alignment is being claimed in the first place? Put another way, if evidence about the strength of a claim is to be included, should it be contained in one of these layers, or in some other place altogether?

I can see the argument being made that credibility-of-alignment is relevant to the goal of improving search relevance (LRMI) and also that endorsement-of-value is an inherent quality of a badge (OBI), but also that such a thing would be infeasibly complex, or a duplication of effort since ranking results is the bread and butter of search engines to begin with.