draft text for UK government

Hi all,

Stella Dextre-Clarke, who contributes to this list, also works on development of the Integrated Public Sector Vocabulary (IPSV), a thesaurus-like subject heading system for use in UK government metadata.  I asked her if IPSV could publish in SKOS Core, and she asked me to draft some text outlining the benefits.  I thought I'd post that here first (Stella I hope you don't mind) so I can get a sanity check, and also some comments and suggestions from you wise folk :)

Recommendations for publication of IPSV to enable use in networked information systems.


Recommendation 1: The [relevant authority] issues URIs (Uniform Resource Identifiers) as the official identifiers for each of the concepts of the IPSV, and for the IPSV itself. 

Explanation: 

Why use URIs?  Quite simply, URIs make your metadata unambiguous, in whatever context it appears.  This is a fundamental enabling step, allowing government metadata to be used alongside metadata from other sources without confusion.  

The use of URIs to identify metadata elements and values is established practice.  For example, the Dublin Core Metadata Initiative has issued URIs for each of the metadata elements in the Dublin Core Element Set and Dublin Core Terms.  The use of URIs is fundamental to the DCMI abstract model, the DCMI recommendation describing a common model for metadata descriptions.  The use of URIs is also fundamental to the Resource Description Framework (RDF), W3C's recommendation for metadata descriptions.

Because the IPSV (I believe) already has non-URI concept identifiers, I would recommend that the URI for each concept be formed by appending the concept identifier to some base URI such as e.g. http://www.esd.org.uk/standards/ipsv/concept/


Recommendation 2: The [relevant authority] makes a formal statement regarding the persistence of URIs issued as identifiers for concepts of the IPSV, and for the IPSV itself.

The DCMI, for example, has made the following statement about URIs used for DCMI metadata terms:

'The DCMI recognizes that people and applications depend on the persistence of formal documents and machine processable schemas that have been made publicly available. In particular, the stability of namespace URIs for metadata terms is critical to interoperability over time. Thus, the wide promulgation of this set of URIs dictates that they be maintained to support legacy applications that have adopted them.'

Explanation:

A statement such as the above explains to the users the degree of confidence they may have in the URIs being used consistently in the future.  

For more on URIs, see 'Cool URIs don't change' <http://www.w3.org/Provider/Style/URI> by Tim Berners-Lee.  Although it describes the use of URIs in the context of their use as identifiers for web documents, the same principles apply to the use of URIs for metadata concepts.

See also the section 'URI Persistence' in the 'Architecture of the World Wide Web, Volume One' <http://www.w3.org/TR/webarch/#URI-persistence>.

If the [relevant authority] doesn't wish to make any commitments to the persistence of issued URIs then it should make a statement to that effect, so that users may have reasonable expectations with respect to the consequences of their use.


Recommendation 3: The [relevant authority] publishes an RDF description of each of the concepts of the IPSV, and of the IPSV, using the SKOS Core Vocabulary as appropriate.

Explanation:

RDF allows metadata descriptions to be distributed, but to still be sensibly combined and integrated as necessary.  This can greatly reduce the cost of maintaining networked metadata applications.  Furthermore, the handling and integration of RDF data can be implemented by standard software toolkits, minimising the need for custom-written software for data integration.

The SKOS Core Vocabulary is an RDF Vocabulary appropriate to the description and publication of thesaurus-like data in RDF.  It is a W3C Working Draft, and although it is still under development, it has a significant early deployment in the instances of the Dewey Decimal Classification, the UK Archival Thesaurus, and the General Multilingual Environmental Thesaurus.



... That should do for the moment.  Anyone have any thoughts on this?

Cheers,

Al.


  
---
Alistair Miles
Research Associate
CCLRC - Rutherford Appleton Laboratory
Building R1 Room 1.60
Fermi Avenue
Chilton
Didcot
Oxfordshire OX11 0QX
United Kingdom
Email:        a.j.miles@rl.ac.uk
Tel: +44 (0)1235 445440

Received on Tuesday, 11 October 2005 18:33:34 UTC