I think the idea of various 'levels' of compliance with different mandatory
elements is a good one, though I would suggest a multipart specification is
better, rather than levels with each a subset of the higher level. For
example there could be a part that mandates annotations and an id.

As several people have pointed out, there is a big difference between having
the information available somewhere in a distributed model and having it in
the instance itself. For many uses the only information available may be in
the instance itself.

The purist view of only mandating the identifiers is fine if it assumes the
id will resolve to more information or a fuller metadata instance elsewhere.
So then why mandate both? The only essential thing in a distributed model
where ids resolve to further information is a meta-metadata identifier that
tells you it exists.

Cheers
Mike 7:-D
-----------
Mike Collett, Schemeta
+44 7798 728 747
------------
www.schemeta.com
email: [log in to unmask]


> From: Pete Johnston <[log in to unmask]>
> Reply-To: Pete Johnston <[log in to unmask]>
> Date: Thu, 12 Aug 2004 15:25:00 +0100
> To: <[log in to unmask]>
> Subject: UK LOM Core: mandatory elements (summary)
>
> [Initial summary from Phil, with some editing/expansion by me!]
>
> *UK LOM Core, mandatory elements (summary)*
> The following is a summary of discussion on the CETIS-Metadata list, and
> will hopefully serve as the agreed starting point for further discussion
> at the UK LOM Core meeting in Glasgow.
>
> The archives of the original discussion can be found at
> http://www.jiscmail.ac.uk/lists/cetis-metadata.html (the discussion took
> place in July).
>
> *Position in UK LOM Core draft 2*
> 18 simple data elements are mandatory:
>
> 1. General: identifier (catalog & entry), title, language, description.
> 2. Lifecycle: contribute (role, entity, date).
> 3. Meta-metadata: identifier (catalog & entry), contribute (role,
> entity, date), schema, language.
> 4. Technical: location.
> 6. Rights: copyright and other restrictions, description.
>
>
> *Issue Description*
> Andy Powell (on Fri 9 Jul at 16:32 subject: UK LOM Core: mandatory
> elements) wrote:
> "My feeling is that UK LOM Core has a major problem with the way it
> mandates so many elements."
> "In summary, conformance with UK LOM Core should tell you more about how
> particular elements and values have been used than about which set of
> elements to expect in a record."
>
> He suggested this arose from a view that metadata records were discrete
> and complete, and that this was not the only view and gave an example
> scenario where this was not the case.
>
> [Lorcan Dempsey expanded on this view by citing a recent paper on
> "Metadata Augmentation" the abstract of which says "The key to this
> augmentation process involves changing the basic metadata unit from
> "record" to "statement."
> http://metamanagement.comm.nsdl.org/Metadata_Augmentation--DC2004.html]
>
> Andy continued:
> "I think that it would be more helpful to allow these services to claim
> compliance with the UK LOM Core, even though they each only contain
> partial information.  It somehow feels wrong, or at least I don't
> understand what we achieve, by saying that the individual services are
> not complient."
> "The other argument against making so many elements mandatory is that
> for all elements (with the possible exceptions of the identifiers) there
> will be some scenarios in which the element has no valid value."
>
> Andy suggested only 1.1 general.identifier and 3.1
> metametadata.identifier should (possibly) be mandatory.
>
>
> There was a lot of debate on this. One issue which arose from this was
> that of "what is the UK LOM Core for". See Phil's separate message
> yesterday, but perhaps worth noting in particular:
>
> Andy Powell (on Wed, 14 Jul  at 16:16, subject Re: UK LOM Core:
> mandatory elements):
> "I still think we need to ask oursleves questions like 'What is the
> purpose of an application profile?', 'What does XML Schema already give
> us that we don't need to replicate in the application profile?', 'What
> does mandatory mean?' and 'What are the benefits and downsides of making
> particular elements mandatory or not?'.
>
> Scott Wilson (on Fri, 16 Jul at 13:18 subject Re: UK LOM Core: mandatory
> elements) wrote that
> "It all comes down to purpose..."
> "Clearly the 4-item and even 18-item sets will not support adaptive
> real-time use, as in typical SCORM/just-in-time training scenarios,
> where objects are discovered and used without necessarily any
> intermediary selecting or modifying the materials. So I wouldn't be able
> to use such objects in that sort of scenario."
> "The more minimal set, like DC, looks more fit for purpose for
> preemptive discovery of materials for hand-assembly into courses, where
> there is a technical expert on hand to establish (by trial and error)
> whether they will "play" in the local environment."
> "Of course, is this a minimal set that should be supported as a search
> index for interoperable search? Or is it disconnected from search
> functions, and is about usage hints to a delivery process? Or is it
> both, everything, or nothing?"
>
>
> On the one hand arguments were made that the UK LOM Core should be
> sufficiently flexible to support a range of practice by metadata
> creators/providers and that its primary purpose should be to encourage
> consistency in the choice of metadata elements and particularly in the
> values provided, in the interests of semantic interoperability.
>
> On the other hand, there was concern that those searching for resources
> were provided with enough metadata to meet their needs.
>
> The summaries below don't represent every contribution, and are grouped
> for / against rather than chronologically (some posts raised points for
> and against).
>
> Andrew Middleton (on Tue, 13 Jul at 11:18 subject Re: UK LOM Core:
> mandatory elements) wrote:
> "A metadata record is best completed by several people. But that means
> records are left hanging (indefinitely?), and so not
> compliant, until all quality inputters have done their bit.
> ...
> At the moment the onus tends to be on one person to do the metadata. As
> a consequence, depending on who does it, some elements are of a high
> quality while others are not and so the metadata record, and purpose,
> suffers."
>
> Fred Riley (on Mon, 12 Jul at 11:39 subject Re: UK LOM Core: mandatory
> elements) wrote that
> -he was worried about how the project he was working on could provide
> all the mandated metadata
> -thought that if there is a problem filling in a mandatory field then
> the result might be nonsense.
>
> Andy Powell (on Wed, 14 Jul  at 16:16, subject Re: UK LOM Core:
> mandatory elements) suggested that it was important to know what
> cataloguing rules were used to provide values for the elements that are
> provided. In summary he wrote:
> "My personal opinion is that the real benefit of something like UK LOM
> Core is in achieving greater (but not 100%) consistency in the choice of
> metadata elements and the way values are constructed for those elements
> across a range of disperate data providers - not about telling consumers
> of metadata records from those services which elements will absolutely,
> definately be available."
> "If my application absolutely needs 4.2 technical.size in order to
> function properly, then I can simply throw any records without that
> element away. I don't need an application profile to enable me to do
> that - I just do it, having looked at each record.  But I do need the
> application profile in order to make sure that everyone constructs the
> value in the same way."
>
> Andy Powell (on Sat, 8 Aug at 22.51, subject: Re: UK LOM Core: what is
> it for (was Re: UK LOM Core: mandatory elements))
> "My concern is that by *mandating* things you may actually end up
> reducing the level of 'meangingful exchange' by forcing people to supply
> a value where none exists, or where the value is unknown."
> "To maximise the level of 'meaningful exchange' we may find it is better
> to
>
> - strongly encourage the use of a particular set of elements
> - strongly encourage (or perhaps even mandate) a particular set of rules
>   for formulating the values of those elements (cataloguing guidelines)
> - but ultimately leave enough flexibility that people can decide when
>   it is appropriate to use an element and when it is inappropriate."
>
>
>
> Phil Barker (on Tue 13 Jul at 15:42, subject Re: UK LOM Core: mandatory
> elements) wrote:
> "We want to provide teachers and learners with enough information to
> locate appropriate learning resources"
>
> John Casey (on Mon 12 Jul at 15:48, subject FW: UK LOM Core: mandatory
> elements) wrote:
> "Andy's ideas for a kind of distributed metadata record (I hope I have
> got that right) has a lot of attraction (especially the idea about the
> annotations service which we are interested in here) but I think for the
> people running large repositories like JORUM etc they will require the
> 'full monty' of metadata at the time of deposit"
>
> Scott Wilson (on Tue 13 Jul at 11:59, subject Re: UK LOM Core: mandatory
> elements) distinguished between permitting partial records within a
> workflow and exposing those exposing them to external systems:
> "If you need to support fragmentary records (that is, with missing
> mandatory data) internally within a workflow, that isn't an
> interoperability issue - but you don't want to be sending such records
> to external targets without prior agreement, or you should expect to
> have those records rejected as being incomplete."
>
> Pete Johnston (on Wed 14 Jul at 13:55, subject Re: UK LOM Core:
> mandatory elements) wrote:
> "The distributed/multi-part approach ...  leaves the service provider
> with the question of how they can be sure that the sum of however many
> parts they gather will provide the minimum data they are depending on to
> provide a service"
>
> Phil Barker (on Tue 13 Jul at 15:42, subject Re: UK LOM Core: mandatory
> elements) introduced the idea of multiple levels of conformance to UK
> LOM Core.
>
> Andrew Middleton (on Wed, 14 Jul  at 15:09, subject Re: UK LOM Core:
> mandatory elements) wrote:
> [Based on WAI levels of compliance] "having 'optional' levels of
> compliance in effect means there is a great tendency amongst academic
> staff and support staff to work to the lowest level."
>
>
> Also, several people commented along the lines of "if it isn't mandatory
> then it won't be filled in", and it was noted that to a software
> application conditions such as "strongly encouraged" and "recommended"
> were the same as "optional".
>
>
> *Routes to Resolution?*
> Clarify what we want to acheive with the UK LOM Core and how we want to
> see it used.
>
> Consider distinguishing between
> - complying with the requirement for elements in the UK LOM Core to be
> present; and
> - complying with the definitions / guidelines for providing values in
> the UK LOM Core
> (we could call the second type of compliance "UK LOM element usage" or
> similar.)
>
> Consider compliance at different levels:
> level 1 compliance could be as Andy suggests with 2 mandatory elements,
> everything else would be highly desirable, desirable or optional;
> level 2 would make the highly desirable elements mandatory (and the
> desirable would be highly desirable?);
> level 3 would be as we have now.
>
> Other suggestions?
>
> 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/