Andy,

I agree with most aspects of your paper. However I do find it hard to
understand how the requirements for an identifier scheme will be decided
without some sort of model that describes what a LO will be and how they
will be used.

I'd like to add the following thoughts:

I imagine LOs to be created and used in a similar way to eJournals. I think
an eJournal is a LO in fact, as are reading lists, even quotes. I expect all
the same issues with eJournal subscriptions to arise when the 'LO economy'
kicks off. Multiple vendors, of different types provisioning with different
levels of service and different tariffs.

With this model I think your paper will cater very well.

One thought is that perhaps LO will be much more dynamic entities than
journal articles. Authors may own a LO such as a discussion room, a
simulation, or a piece of software and want to be constantly re-releasing
it. With each release they may want to change the tariff so may well have to
assign a new ID.

A question is what functionality will LOM repositories like the Learning
Object Network, Jorum+, Universal, and software like DSpace or Eprints
provide. What functionality will be provided for LO maintenance in terms of
pricing, release maintenance, relationships between LOs.

The need for a GUID is essential for institutions to not be tied into
provisioning a LO from a single Target, i.e. using their Resolver to find
the Appropriate Copy of a LO. I just think we need to consider scenarios
where the LO is more dynamic than the eJournal model for provisioning
journal articles.

Do we need to consider re-allocation of IDs into a pool.
How will links between metadata and LO instance be maintained?

Howard


I am very interested to see what comes of your meeting on the 1st.

-----Original Message-----
From: Phil Barker [mailto:[log in to unmask]]
Sent: 30 April 2003 14:40
To: [log in to unmask]
Subject: Re: Identifiers for learning objects - a discussion paper

Lorna M. Campbell wrote:
> Hi Andy,
>
> Thanks very much for your discussion paper.  Having all the issues
> relating to identifiers outlined so clearly is a _big_ step forward. I'm
> cc'ing this mail to the EC SIG mailing list too as many of the issues
> will be of direct relevance to members of that group too.
>
> In terms of comments, there's only one issues I would quibble with....
>
>> 6. Simple to assign: The process of assigning identifiers to learning
>> objects should be as simple as possible. Assignment should be
>> independent of the workflow associated with creating and managing a
>> learning object. In particular, assignment should be independent of
>> the process of depositing the learning object in a repository. For
>> example, if a person creates a learning object on their PC, and
>> packages it using a desktop tool, creating metadata about it in the
>> process, then they should be able to assign an identifier to both the
>> learning object and the metadata record and have those two identifiers
>> honoured (i.e. used) by a repository (or several repositories) when
>> the learning object is deposited. The creator doesn't want the
>> repository to assign a different identifier to the object and metadata
>> when they deposit them: nor do they want to have to wait until the
>> point of deposit before they can assign an identifier.
>
>
>
> I'm not so sure about this.  While I agree that in theory that
> assignment should be independent of workflow I don't know whether this
> is really practical.  Although the primary aim of the UKCMF is to
> facilitate interoperability, a no less important goal is to simplify the
> creation of metadata as much as possible.  Part of this simplification
> will probably (hopefully?) involve repositories generating as many of
> the UKCMF fields as possible.  In the back of my mind I had envisaged
> that identifiers would also be generated automatically but I have to
> confess I hadn't thought through the practicalities and implications of
> this approach.  Still I can't help wondering how many authors are likely
> to add identifiers to resources and metadata at the point of creation
> rather than assuming that the database or repository will do it for them
> at the point of upload?
>

But what if the creator uploads the object to two repositories which assign
two different identifiers? Nothing wrong with that, until a third
repository (or resolving service, or cross searching portal) gets hold of
[metadata for] both copies and needs to work out whether they are
duplicates [and it will need to work out whether what it has are two
different metadata records for one resource, or duplicate metadata records].

There will be ways round this problem, but I think that it would still be
best to avoid unecessarily multiplicity of object ids. So I would suggest
that ids need to be allocated as early as possible (ie during the creation
or the initial publishing of the object) and repositories which receive an
object with an existing id should not create a new id--or (more
realistically) if the repository does create a new id, it should keep it to
itself.

Phil


--
Phil Barker                            Learning Technology Adviser
      ICBL, School of Mathematical and Computer Sciences
      Mountbatten Building, Heriot-Watt University,
      Edinburgh, EH14 4AS
      Tel: 0131 451 3278    Fax: 0131 451 3327
      Web: http://www.icbl.hw.ac.uk/~philb/