Here is an example that demonstrates how Schema.org vocabulary terms can
be mixed with other RDF vocabulary terms to describe various things in
RDF/XML (no HTML "carrier" needed):
http://rdf-translator.appspot.com/convert/rdfa/xml/html/http%3A%2F%2Fwww
.worldcat.org%2Foclc%2F136259
The Schema.org properties are all in the Schema.org namespace because
RDF requires namespaces for all vocabulary terms. They are merely
playing by the RDF rules like everyone else.
They could have reused some existing vocabulary terms like dct:creator
instead of coining their own schema:creator, but this is true for any
RDF vocabulary. Reuse isn't required. Search engines have enough clout
to make such mappings our problem rather than theirs. Again, this is the
right of any RDF vocabulary publisher. The fact that we don't have as
much clout as they do is our problem, not theirs.
I do not think Schema.org is the only way to create linked data and I
only consider it the "best" in the senses that a) search engines can
consume and (re)use it in valuable ways, b) it is useful and
understandable across industries, c) it's pretty intuitive and thus
relatively future-proof, d) etc. Will something come along in 5 years to
replace Schema.org? I don't know, but I'm not terribly concerned because
of c) above.
As before I vehemently deny the statement that "it truly is a microdata
format". It is not. It is a vocabulary that can be serialized as
Microdata. There is a difference.
In order for libraries to host a vocabulary as good and useful as
Schema.org or FOAF we would have to invent one first. If it turns out to
be BIBFRAME, I will be thrilled. In the mean time, I'm sure there will
be enough copies of the Schema.org and FOAF ontologies laying around
after the apocalypse that we will be able to muddle through. :-)
Jeff
-----Original Message-----
From: Bibliographic Framework Transition Initiative Forum
[mailto:[log in to unmask]] On Behalf Of Karen Coyle
Sent: Thursday, March 14, 2013 2:06 PM
To: [log in to unmask]
Subject: Re: [BIBFRAME] Reuse (or not) of existing ontologies
On 3/14/13 10:29 AM, Young,Jeff (OR) wrote:
> Schema.org is NOT "yet another monolithic vocabulary". It also isn't
> dependent on HTML markup.
Well, the schema.org properties are all in the schema.org namespace, so
there is http://schema.org/title, not http://purl.org/dc/title, or
http://rdvocab.info/Elements/Title. Obviously you can put whatever
metadata you want into your HTML file, but schema.org is schema.org. I'm
quite unclear as to why you think this is the only or best way to create
linked data, since it truly is a microdata format designed for HTML
mark-up.
And speaking of schema.org, the namespace is owned by Google, which to
my mind makes it essentially proprietary. Yes, all namespaces need to be
owned by someone, but I would prefer to use properties that are backed
by institutions with the kind of longevity that we in libraries aim to,
and often do, achieve. This has been my criticism of FOAF -- a fine
vocabulary but no solid institutional backing. There have been
discussions of using library archiving skills and facilities to archive
key vocabularies, and FOAF seems to me to be a contender. This could
bring something like FOAF into library space (if we do want to use FOAF
properties, of course). What we lack in most vocabularies is good
versioning so that a vocabulary can evolve without breaking current and
past usage. But that's a whole 'nother can of worms.
kc
kc
> It IS inherently Linked Data.
>
> Jeff
>
> -----Original Message-----
> From: Bibliographic Framework Transition Initiative Forum
> [mailto:[log in to unmask]] On Behalf Of Karen Coyle
> Sent: Thursday, March 14, 2013 10:54 AM
> To: [log in to unmask]
> Subject: Re: [BIBFRAME] Reuse (or not) of existing ontologies
>
> On 3/14/13 3:43 AM, Owen Stephens wrote:
>> On 13 Mar 2013, at 21:13, Eric Miller <[log in to unmask]> wrote:
>>> Our initial linked data analysis identified literally dozens of
>>> namespaces used in the Library community reflecting various stages
>>> of experimentation. More if you start to broaden analysis to the
>>> include museums, archives, galleries, etc.. And even more if you
>>> broaden into more of how we might imagine patrons / others adding
>>> value to this data. This shouldn't be a surprise to anyone familiar
>>> with RDF / Linked Data. There isn't one vocabulary that works (or
>>> will work) for everything. ;)
>> I'd absolutely agree with this - which is perhaps why the idea of a
> 'bibframe' vocabulary worries me. While I absolutely see the point of
> a framework in which libraries operate to describe resources,
> approaching the description of resources from a library perspective
> only worries me more. Even if we are creating new vocabularies (for
> the reasons already put forward), I wonder if there is an argument for
> having a more 'small bits' approach to building these than trying to
> create a single vocabulary to tackle the whole of library description?
>
> First, I want to note that RDA, the cataloging rules that prompted the
> idea that we need to move away from MARC to something more akin to
> linked data, already has an RDF vocabulary and its own namespace. [1]
> This is managed by JSC. So either BIBFRAME will duplicate the well
> over
> 1,000 RDA properties, or it must make use of that property set. It
> isn't
>
> clear why BIBFRAME isn't using the RDA RDF where it applies.
>
> Second, I'm seeing some confusion about what BIBFRAME is. Is it a
> framework? Is it a framework + vocabulary? I was under the impression
> that BIBFRAME was providing a basic model of "things in the
> bibliographic space" within which a variety of vocabularies could be
> used. What concerns me about the current direction is the emphasis on
> converting MARC records. At least experimentation with RDA would be
> looking toward the future rather than the past.
>
> And finally, my suggestion about using a "public facing" form of our
> data is not the same as Richard's suggestion of schema.org. While it
> may
>
> become well-known, schema.org is "yet another monolithic vocabulary."
> It
>
> is also wrapped around displays in HTML. While it may evolve into
> something else, it isn't inherently linked data. We need to be linking
> to data in the LOD space ASAP if we want to be "of the web." There are
> many possibilities for that, but we aren't taking advantage of them.
>
> kc
>
> [1] http://rdvocab.info
>
>> If this was done carefully we may even encourage other communities to
> take up some of the family of vocabularies used within bibframe which,
> to be honest, seems unlikely if what we end up with is a big 'library
> vocabulary'. It might also say to the rest of the community that we
> ('libraries') understand that this isn't about creating 'one vocab to
> rule them all' :)
>> It may be that some of the concerns I have can be mitigated by simply
> declaring equivalence between bibframe entities and entities described
> by others using other vocabularies. I also acknowledge Richard's point
> that 'bibframe' may not be what is typically exposed to the outside
> world. However I think there is a high risk that at this stage
> bibframe is seen as being too insular - I'm not sure if there is a
> solution to this, but I'm concerned we may lose support of those who
> have already invested time and effort in building vocabularies and
> describing things that libraries are interested in (both inside and
> outside the library
> community) - which I would argue are the very communities we should be
> seeking to engage.
>> Owen
>>
>> Owen Stephens
>> Owen Stephens Consulting
>> Web: http://www.ostephens.com
>> Email: [log in to unmask]
>> Telephone: 0121 288 6936
--
Karen Coyle
[log in to unmask] http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet
|