As a general observation about ontologies, by and large they are open
to the public for use and extension. So if the vocabulary is
represented publicly as an ontology, that would be an outcome. (A
positive outcome in my opinion.) Other representations could
simultaneously be provided, as by Roy's service for example.
In many communities, the provision of ontologies for semantic
information does not carry any additional burdens with regard to
validity, consistency, or (most important) persistence and versioning.
It is the opinion of many in the semantic community that persistence
and versioning are irrelevant at best, or even counter to semantic web
goals. Long story. However, many of us believe that these additional
burdens are fundamental to the way the science community needs to
store, use, and access this information.
There are at least two services available for marine and related
vocabularies today that promote the latter approach. Roy's NERC
service (operational for several years, and much longer including
previous incarnations) and MMI's vocabulary service (in operational
alpha/beta for about a year now) both invest many of their resources
to encourage best practices for managing vocabularies [1]. I expect
either of our organizations would be happy to have discussions with
any community that wants to know about, or pursue, the establishment
of 'authoritative practices'. We would also be pleased (relatively!)
to hear about places we are falling short of the mark.
As to URIs, I concur with Roy that development of those services --
vocabulary storage, versioning, URI resolution, and so on -- is a bit
of a hill to climb for any organization, and if the services provided
by the existing organizations are not sufficient, engagement to
improve them is perhaps the best strategy, and almost certainly the
cheapest.
On the other hand, there will be local decisions to make about the
form of URI to use (URN, URL, something else), and whether to use
versioned forms or unversioned forms. These are within the domain of
CF to make, and have been discussed a bit previously on this and other
CF threads. But fair warning -- these topics have been discussed at
great length in many organizations, without obviously correct answers
emerging. I suggest you kick that topic over to the TRAC system, and/
or other CF authority, if you really want to have a go.
John
[1] MMI Semantic Framework information:
http://marinemetadata.org/semanticframework
On Oct 28, 2009, at 0654, olivier lauret wrote:
> Hi Roy,
>
> This discussion has reached some very interesting points to come in
> the near future.
>
> If we introduce such extents related to CF, using URIs, personally I
> am not totally against it and I believe this example with flags is
> an excellent illustration of that need.
> And probably we'll need " dedicated organisation with a high level
> of professionalism" like yours, Roy :) But I am also persuaded that
> ontologies (it is indirectly that topic) are not forced to obey
> services or be under control from a service provider. The philosophy
> of ontologies, and technologies like OWL, is carrying some knowledge
> and enrich this knowledge by sharing it. Maybe an authority is
> needed to control and ensure the integrity of a vocabulary, like a
> database administrator. But everyone should be free to use the
> source file and enrich/adapt it to its own needs. I sincerely hope
> this will happen. (end of the disgression)
>
> Cheers
>
> Olivier.
> -----Message d'origine-----
> De : cf-metadata-bounces at cgd.ucar.edu [mailto:cf-metadata-bounces at cgd.ucar.edu
> ] De la part de Lowry, Roy K
> Envoy? : mercredi 28 octobre 2009 11:52
> ? : Heinke Hoeck; V. Balaji; cf-metadata at cgd.ucar.edu
> Objet : Re: [CF-metadata] Dealing with large numbers of flag
> valuesinnetcdf cf
>
> Hi Heinke,
>
> I am involved in this issue through an IODE/SCOR group on data
> publication. I see the development of identifier allocation and URI
> resolution as outside the scope of CF. It is something that needs
> to be provided as a service by a dedicated organisation with a high
> level of professionalism. If existing service providers fall short
> of the mark then this should be documented and taken up with the
> service providers. IODE, SCOR and possibly CF governance are
> appropriate vehicles for this.
>
> Your e-mail triggered the thought in my mind that now we have
> versioning fully implemented on the Vocabulary Server bringing it up
> to publication standard there is the possibility of assigning a DOI
> to each published version of a vocabulary. I'll give this some
> thought.
>
> Cheers, Roy.
>
> -----Original Message-----
> From: cf-metadata-bounces at cgd.ucar.edu [mailto:cf-metadata-bounces at cgd.ucar.edu
> ] On Behalf Of Heinke Hoeck
> Sent: 28 October 2009 10:40
> To: V. Balaji; cf-metadata at cgd.ucar.edu
> Subject: Re: [CF-metadata] Dealing with large numbers of flag
> valuesinnetcdf cf
>
> V. Balaji wrote:
>> Bryan Lawrence writes:
>>
>>> My job is to persist data indefinitely, and I'm relying on
>>> vocabulary
>>> servers ... (and in particular, Roy's ... ) somehow we will persist
>>> the vocab.ndg.nerc.ac.uk address, regardless of nerc's future, and
>>> of
>>> ndg's future, although if ac.uk went away we might be in
>>> trouble :-).
>>
>> Could happen...
>>
>> http://www.guardian.co.uk/politics/2009/oct/18/conservatives-defence
>>
>> I see nerc.co.uk coming...
>
> I would like to discuss the use of persistent identifiers.
> Presently, We are using the DOI and URN persistent identifiers
> for the publication and reference of primary data (www.std-doi.de).
>
> For our application, of course, the DOI (http://dx.doi.org/) and URN
> (http://www.persistent-identifier.de/resolving/ResolverDemo.php)
> systems are not adequate. However, the idea of persistent identifiers
> or keywords could solve the problem of the persistent metadata
> identification in the NetCDF-CF header. The persistent identifier
> doesn't need to contain semantic meaning. The URL before it could be
> changed but the object (meaning) behind must be persistent.
>
> We wold need a resolver system which should be installed at more than
> one institute. I am not an expert to set up a resolver system but
> may be there is anyone who has experience with it.
>
> Best wishes
> Heinke
>
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
>
> --
> This message (and any attachments) is for the recipient only. NERC
> is subject to the Freedom of Information Act 2000 and the contents
> of this email and any reply you make may be disclosed by NERC unless
> it is exempt from release under the Act. Any material supplied to
> NERC may be stored in an electronic records management system.
>
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
>
>
> Cliquez sur l'url suivante
> https://www.mailcontrol.com/sr/DJXSJs6NHaPTndxI!oX7UlEvFrFJ1EkC60glhBAlxG0IwiY5afkpLFUYRifLIyv!o1Oe2QXkTzyDEGNPdlyNkQ==
> si ce message est ind?sirable (pourriel).
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
---------------
John Graybeal
Marine Metadata Interoperability Project:
http://marinemetadata.org
graybeal at marinemetadata.org
Received on Wed Oct 28 2009 - 18:28:55 GMT