On Mon, Aug 16, 2010 at 09:55:16AM +0100, Andy Powell wrote:
> I agree this is interesting. I guess this is quite likely to become the norm (if it isn't already).
>
> On that basis, I suppose one could argue that a success
> criteria for DCMI will lie in the use of our ontology as a
> kind of secondary 'equivalence' or 'switching' layer between
> other ontologies, rather than (or as well as) in its direct
> use in instance metadata ?
At [1], Felix has just explained that the "ma:" namespace is
not designed primarily to be used in annotating resources,
but to provide precise mappings, via an API, to elements in
widely-used formats:
The main application scenario for the ontology is access to existing
multimedia metadata . Now, if you take an example like XMP
http://www.w3.org/TR/mediaont-10/#d0e9670 , there is a title in terms of
dc:title, and a title in terms of xmpDM:album. One has language information
associated, the other hasn't. One is defined as an array, the other isn't.
For a JavaScript developer who wants to use the media annotation ontology
mappings via the API for media resources
http://www.w3.org/TR/2009/WD-mediaont-api-1.0-20091020/ , this mapping
information is crucial. Note that such constraints are rather technical
(that is "data type" constraints), not a question of control about external
vocabularies.
There is another aspect: the only application scenario of the ontology is:
being a mediator between existing formats, and the main usage scenario is to
get information about a media resource, but not to annotate a resource using
the "ma" properties. That puts the properties on an orthogonal level,
compared to dublin core or other formats, who are mostly used for direct
reading and writing. Note btw., that there was a discussion about a
requirement to write metadata via the ontology
http://www.w3.org/TR/2010/WD-media-annot-reqs-20100121/#req-r02 , but that
so far did not lead to any results, since most of the mappings lead to loss
of information, so writing in the base format does not work.
Tom
[1] http://lists.w3.org/Archives/Public/public-xg-lld/2010Aug/0072.html
>
> Andy
>
> --
> Andy Powell
> Research Programme Director
> Eduserv
> t: 01225 474319
> m: 07989 476710
> twitter: @andypowe11
> blog: efoundations.typepad.com
>
> www.eduserv.org.uk
>
> -----Original Message-----
> From: DCMI Architecture Forum [mailto:[log in to unmask]] On Behalf Of Thomas Baker
> Sent: 14 August 2010 16:32
> To: [log in to unmask]
> Subject: Dublin Core in W3C Ontology for Media Resource
>
> I note with interest, in "Ontology for Media Resource 1.0,
> W3C Working Draft 08 June 2010" [1] that, while strongly anchored
> in the Dublin Core vocabulary, the specification defines its
> own properties with "mappings" to DCMI metadata terms, with the
> following rationale:
>
> The ontology defines mappings between a set of vocabularies
> and a set of core properties in our own namespace, which is
> identified with the "ma" prefix in this document. Although
> some of these properties can appear to be redundant with
> the Dublin Core set of properties, the set of properties
> that make up our ontology are defined in a new namespace
> that is separate from the Dublin Core for several reasons,
> including:
>
> * Dublin Core is only one of the vocabularies that mappings
> are defined for.
>
> * The Dublin Core set vocabulary does not cover all of our
> needs; hence, we would still have to create new properties
> in our own namespace.
>
> * More importantly, the Dublin Core properties have
> been created with a set of restrictions. While these
> restrictions are in general somewhat loose, we may want
> to apply other restrictions to our properties. Therefore,
> we have to define our own set of properties that we can
> control (e.g., by constraining their allowed values);
> hence, these properties cannot be dependent on any one
> (or even several) external source(s) of authority for
> the definition of our core mapping.
>
> For a practical use of the media ontology in an API,
> we define type restrictions for our properties that go
> beyond the generic Dublin Core specification.
>
> Some relevant mappings can be found in 4.2.2.3 (EBUCore),
> 4.2.2.8 (MediaRDF), and 4.2.2.16 (XMP) [2,3,4].
>
> As Felix Sasaki explains in [5], the mappings in [1] are
> designed for converting between fixed formats (EBUCore,
> MediaRDF, XMP, etc) using API methods defined in [6],
> but an RDF-based ontology (possibly with formally defined
> mappings?) is on the way.
>
> Tom
>
> [1] http://www.w3.org/TR/2010/WD-mediaont-10-20100608
> [2] http://www.w3.org/TR/mediaont-10/#d0e5511
> [3] http://www.w3.org/TR/mediaont-10/#d0e3064
> [4] http://www.w3.org/TR/mediaont-10/#d0e9670
> [5] http://lists.w3.org/Archives/Public/public-xg-lld/2010Jul/0075.html
> [6] http://www.w3.org/TR/2010/WD-mediaont-api-1.0-20100608/
>
> --
> Thomas Baker <[log in to unmask]>
--
Thomas Baker <[log in to unmask]>
