Template talk:Term

This is a discussion page. To start a new topic, create a header (enclosed by "=="), to respond use indentation (":" in first line). Please sign all entries with date and username; adding ~~~~ will trigger the Wiki to do this for you.

Todo

Find concise property names for “Part of” and “Is a” to get a suitable auto completion for users --Andreas Plank 14:46, 19 May 2011 (CEST)

Property:DCMI:isPartOf for now. --Andreas Plank 17:59, 22 September 2011 (CEST)

Mapping

Proper mapping in SMW to get correct RDF

Mapping relations have usually a “from to”-relation. So what should be mapped to exactly (per default)? The “from to”-relation is not implemented. --Andreas Plank 14:19, 22 September 2011 (CEST)

Resolving vocabularName, identifiers, subcollections properly

Problem: resolve vocabularName (subcollections) for GBIF, e.g. taxon_rank, to a correct rdf:about="…" in owl:ObjectProperty or swivt:Subject and how should we construct wiki page names then? Because the rdf:about="…" is constructed by the value in [[imported from::…]]. Take the XML-example of taxon_rank/speciesAggregate from GBIF:

 page name as GBIF:species aggregate or GBIF:speciesAggregate …
 │  or GBIF:taxon rank/species aggregate or GBIF:taxon rank:species aggregate?
 │   ┌ label= species aggregate
 │   ├ code = speciesAggregate
 │   ├ collection = GBIF:taxon_rank …
 ↑   ↑   └ or: collection= GBIF + subcollection= taxon_rank?
 -------- at GBIF -------------- 
 speciesAggregate (dc:identifier)
 Species Aggregate (preferred term dc:title)
 speciesAggregate (concept dc:title)
 -------- at GBIF -------------- 
 ↓
 RDF export
 └ [[imported from::GBIF:taxon_rank/speciesAggregate]] 
                         └────────────┬─────────────┘
                        must become a rdf:about of:
   http://vocabularies.gbif.org/services/gbif/taxon_rank/speciesAggregate

Proposal: use subpages and an additional template parameter “subcollection” if there is an original sub (vocabulary) collection that is reused here. Page names should be in natural language, where possible, so no GBIF:taxonRank/speciesAggregate-something. --Andreas Plank 17:59, 22 September 2011 (CEST)

I cannot completely follow all you intend here. I only make some observations:

1. identifier given as dc:URI seems to be missing from analysis
2. With respect to "http://vocabularies.gbif.org/services/gbif/taxon_rank/speciesAggregate"

we need to ask GBIF whether they intend to invent new URIs for the tdwg terms, or whether this is a bug.

1. Our "collection" concept is probably matching what the gbif vocab server calls a vocabularly. My main problem is that the place where this is expressed seems to be in dc:URI (see below) which looks inappropriate.
2. About "Page names should be in natural language, where possible" - in principle I agree. So, as far as I understand, we can make our own, internal synonym and still export the RDF of everything pointing to that with the official URI? However, gbif vocab server has the natural language strings under the lists of terms, one for each language. In the wiki this creates a problem: do we create on page in each language? Do they redirect to a single one? Will this single one then be based on the identifier code like GBIF: taxon_rank/speciesAggregate?
3. I post my own tenative annotations/questions to an example record below

Gregor Hagedorn 22:28, 22 September 2011 (CEST)

Annotated vocabularies.gbif.org example

The example is from http://vocabularies.gbif.org/services/gbif/taxon_rank/cultivarGroup with questions by me: Gregor Hagedorn 22:36, 22 September 2011 (CEST)

<concept dc:URI="http://rs.gbif.org/terms/1.0/taxonRank"
  NOTE GH: This seems to be the URI of the collection/entire
  vocabulary, not the URI of the concept. We think this is a bug.
  For information: http://rs.gbif.org/terms/1.0/taxonRank is only a URN
  which presently does not resolve (which is not necessary of course).
  Also note that the web interface displays: 
    http://rs.tdwg.org/ontology/voc/TaxonRank#Cultivar-Group
  The above seems to define a new URI for the tdwg term, 
  without giving the relation anywhere

dc:created="2009-08-03T15:59:53+01:00"
dc:description="http://vocabularies.gbif.org/taxon_rank/cultivarGroup"
  NOTE GH: Using dc:description for the URI of the concept seems
 odd and not within the usual concept of Dublin Core
 (http://www.dublincore.org/documents/dcmi-terms/#terms-description )
 It is clear, that the term has multipe URIs: one on tdwg, one on rs.gbif, 
 one on http://vocabularies.gbif.org. But I think this is not 
 transparent solution.

dc:identifier="cultivarGroup" 
  NOTE GH: The identifier here does not bear any prefix / namespace that 
  makes it anywhere likely to be unique. It is just a literal string. While
  this is in line with dc:identifier, I wonder whether this is intended.

dc:modified="2009-08-03T15:59:53+01:00"
dc:title="cultivarGroup" notes="" reference="">
<preferred>
  <term dc:created="2009-08-03T15:59:53+01:00"
    dc:modified="2009-08-03T15:59:53+01:00"
    dc:title="Cultivar-Group" reference="en"
    standardConcept="cultivarGroup" xml:lang="en"/>
  NOTE GH: It is not yet clear to me what the semantics of "standardConcept" is. Since this is nested within concept, a backlink to a concept seems to be unnecessary.
</preferred>
<alternative/>
</concept>

GBIF has a “vocabularyName” (for the webservice) or “Vocabulary” on the vocabulary overview. --Andreas Plank 10:19, 23 September 2011 (CEST)

I believe the vocabularyName is similar to our collection Gregor Hagedorn 18:29, 23 September 2011 (CEST)

The question remains: How should we deal with that for a page name and the template: term field collection?

• page name like <collection>:<subcollection>/<term>, e.g. GBIF:taxon rank/species aggregate (I think that's safe if term might be defined twice, which is unlikly, I know)
• page name like <collection>:<term> GBIF:species aggregate. Assuming of course the term is unique within <collection>. <subcollection> or vocabulary or what ever we call it should then be saved to SMW

So we must decide for:

1. page name → <collection>:<subcollection>/<term>
   1. additional “subcollection”-template field or
   2. no additional field, but let the user just type for a collection “<collection>:<subcollection>”, e.g. “GBIF:taxon rank”
2. page name → <collection>:<term>
   1. additional “subcollection”-template field or
   2. no additional field, but let the user just type for a collection “<collection>:<subcollection>”, e.g. “GBIF:taxon rank”

For simplicity I'd prefer: 1.2. There are already enoug form fields. Andreas Plank 10:21, 26 September 2011 (CEST)

GBIF Vocabulary (prototype)

The GBIF Vocabulary is still only a prototype under evaluation. Most of the content is added to test the functionality of the tool and not quality checked. The examples above are from the taxonRank vocabulary. I believe it would be more illustrative to explore the terms from the rank vocabulary - while keeping in mind that the version in use by the GBIF infrastructure is available from the TDWG Ontology and from the GBIF registry. Dag Endresen 17:22, 23 September 2011 (CEST)

Note: Markus Döring explains that rs.tdwg.org is the main GBIF endpoint, and is manually maintained as static files in google code. Gregor Hagedorn 18:29, 23 September 2011 (CEST)

Page names of terms

I discussed with Gregor and we both agree that, beside all this technical stuff (sub collection etc.), it is a good idea to keep page names as simple and human readable as possible, because an end user should not know if he must use [[GBIF:taxon rank/species]] to link to a definition of the term “species” etc. The only question remains then, how this can be inserted into the template: term and form:term to be simple for a user as well. --Andreas Plank 18:59, 26 September 2011 (CEST)

Page transclusions

I propose a transclusion of term definition pages on normal pages (see species). It can be done by {{:term definition page}}. The question is what the RDF export looks like. --Andreas Plank 17:30, 28 September 2011 (CEST)