I suppose if people want to be unambiguous, they should just use URIs. 

Exactly that.

I know this discussion is mainly focussing on RDF and such like, but the DwC-A format was designed to make vocabularies a little more accessible for people using spreadsheets / databases etc.

In a data file, you might have a CSV with:

1, Puma concolor, male

And in the meta.xml file, you could have the following, which declares the vocabulary used for the :

  <core encoding="utf-8" fieldsTerminatedBy="," linesTerminatedBy="\n" rowType="http://rs.tdwg.org/dwc/terms/Occurrence">
    <files>
      <location>occurrence.txt</location>
    </files>
    <id index="0" />
    <field index="1" term="http://rs.tdwg.org/dwc/terms/occurrenceID"/><archive xmlns="http://rs.tdwg.org/dwc/text/" metadata="eml.xml">
    <field index="2" term="http://rs.tdwg.org/dwc/terms/scientificName"/>
    <field index="3" term="http://rs.tdwg.org/dwc/terms/sex" vocabulary="http://rs.gbif.org/vocabulary/gbif/sex.xml"/>
  </core>

This is a primitive example, but keeping content and mapping somewhat separate helps keeps this accessible for wider audiences without having to rewrite content every time standards, namespaces, versions etc all change.  The data will simply get outdated quickly as I strongly suspect most content is immutable once written once, most data mappings are immutable once created once and most application developers do not use RDF aware technologies (why I originally went for XML and not RDF in DwC-A meta definition despite DwC itself being in RDF). 

I complete agree that we need to provide the means to declare (strongly versioned) vocabularies, but I tend to suspect that they should be separated from the content for the most part for better reusability - that was the rationale when I introduced it to the DwC-A.  Perhaps it is time to consider a review of the DwC-A meta format though, to play better with (e.g.) SKOS, but I don't think I understand the versioning aspects there - everything seems in constant flux to me.

Cheers,
Tim
 


On Oct 3, 2013, at 2:25 PM, Hilmar Lapp wrote:


On Oct 3, 2013, at 6:42 AM, Steve Baskauf wrote:

I suppose if people want to be unambiguous, they should just use URIs. 

Exactly that.

Unfortunately, I'm not aware of URIs for the terms from various systems of describing biomes.

To the extent that unawareness is to be supported, this can be solved through better documentation.

  I suppose someone could mint terms if they cared about this.

Certainly not - minting random terms whose semantic content consists solely of a non-resolving URI string or a string in whatever natural language is barely better than providing no value at all. 

Or use ENVO uris.

Yes, indeed. If they don't exist yet, ask that they do. That's the way to build community ontologies - by a community actively working to make the ontology what it needs it to be.

-hilmar
-- 
===========================================================
: Hilmar Lapp  -:- Durham, NC -:- informatics.nescent.org :
===========================================================



_______________________________________________
tdwg-content mailing list
tdwg-content@lists.tdwg.org
http://lists.tdwg.org/mailman/listinfo/tdwg-content