1. The NERC DataGrid Vocabulary Server: an operational system with distributed ontology potential Roy Lowry British Oceanographic Data Centre GO-ESSP 2008, Seattle, September 2008

2. Presentation Summary  What we’ve got  How to use it  What’s wrong with it?  What we’re doing about it  What’s in it for CF?

3. What We’ve Got  Server Contents (2008-08-21)  112 public lists  122603 concepts  78123 mappings (RDF triples)  Server Usage 2008 (to 2008-08-21)  2233803 total hits (2000000 of these attributable to robots)  37462 vocabulary catalogue hits  50458 vocabulary list downloads  2085 vocabulary mapping queries  In summary a heavily populated, heavily used system

4. How to Use It  There are four ways to access vocabulary content  Term and list URLs  HTTP-POX web service calls  SOAP web service calls  Interface clients

5. How to Use It  Term and list URLs  Any list may be accessed using a URL of the form http://vocab.ndg.nerc.ac.uk/list/list_id/list_version  Delivers the list, without mappings, as a SKOS XML document  Any term my be accessed using a URL of the form http://vocab.ndg.nerc.ac.uk/term/list_id/list_version/term_id  Delivers the term, including all mappings, as a SKOS XML document

6. How to Use It - SDN:P021:25:PHYC Phycobolin pigment concentrations in the water column WC_PhycobolPig Concentration of phycobolin group pigments such as phycocyanin and phycoerythrin in the water column 2008-03-11T11:56:27.531+0000

7. How to Use It  HTTP-POX service calls  Any API method may be invoked using an HTTP get call  Lists and terms specified in the get call parameters as URLs  Delivers an appropriate XML document (BODC- designed schema)  Documentation at http://www.bodc.ac.uk/products/web_services/vocab/methods.html http://www.bodc.ac.uk/products/web_services/vocab/methods.html  SOAP web service calls  WSDL may be found at http://vocab.ndg.nerc.ac.uk/http://vocab.ndg.nerc.ac.uk/  Same output and documentation as HTTP-POX

8. How to Use It  Interface clients  Maris client set up for SeaDataNet at http://seadatanet.maris2.nl/v_bodc_vocab/welcome.aspx http://seadatanet.maris2.nl/v_bodc_vocab/welcome.aspx  BODC clients at http://vocab.ndg.nerc.ac.uk/ cover more vocabularies if interests extend beyond SeaDataNet http://vocab.ndg.nerc.ac.uk/

9. How to Use It  Typical Maris client output

10. How to Use It  In SeaDataNet vocabulary terms are included in data and metadata as URNs  These have the form:  SDN:list_id:list_version:term_id, e.g.  SDN:P021:23:TEMP  SDN:P021::TEMP (for current version)  URNs resolve to URLs by simple string substitution (service to do this planned)  URLs may be used to obtain the term identifiers, labels and mappings in an XML document

11. What’s Wrong With It?  Historic version serving not implemented  Current version served whatever version is requested  Predicates (SKOS mappings) semantically limited  More suited to a thesaurus rather than an ontology  Richer set in the triple store, but cannot be served without WSDL changes

12. What’s Wrong With It?  Vocabularies not labelled with content governance authority  Mappings restricted to concepts within the server  If a vocabulary is to be included in a mapping then it must be loaded in the server  Not exactly Semantic Web!

13. What We’re Doing About It  Historic version serving not implemented  Necessary SQL now written so functionality will be incorporated into the next release  Predicates (SKOS mappings) semantically limited  Collaboration with Sean Bechhofer (Manchester Computer Science) to design better SKOS documents for server payload

14. What We’re Doing About It  Vocabularies not labelled with content governance authority  On the list of things to do for version 1.2  Mappings restricted to concepts within the server  Developing interoperable server in collaboration with ICES  Triple store structural redesign under consideration

15. What We’re Doing About It  In practice  V1.2 planned for the end of the year  Historic version serving  Vocabularies labelled with content governance  V2.0 date not yet fixed  Linkage to external resources  Improved SKOS document payload

16. What’s in it for CF?  Currently serve Standard Names and Cell Methods  Synchronised with CF site (Alison provides update material)  Mappings maintained to SeaDataNet (BODC) and GCMD vocabularies  Support for representation of Standard Names as URLs that may be mapped and used as URNs

17. That’s All Folks Thank you for your Attention Any Questions?