-
Notifications
You must be signed in to change notification settings - Fork 7
Teleconference 7 8 May 2019
- May 7, 9 pm UTC meeting: Steve Baskauf, Anne Fuchs, Jeff Gerbracht, Niels Klazenga, Cam Webb, Greg Whitbread. May 8, 8 am UTC meeting: Anne Fuchs, Niels Klazenga, Greg Whitbread
After comments from @stanblum at the previous meeting, we discussed the Darwin Core terms we have adopted so far (in general) and decided to use terms in the TNU (working name) namespace instead. The reasons for this are:
- Some of the terms have slightly different meaning in Darwin Core than they will have in TNU.
- While almost everything we have discussed so far can be done with Darwin Core, the Darwin Core Taxon class is used for a lot of other things as well and one can't see from a Darwin Core Taxon core data exactly what sort of data it contains and therefore it is difficult to interpret the data or assess its fitness for a particular purpose. Because of this, we think it's better to have TNU for the sole purpose of transferring taxonomic data in an unambiguous way and the Darwin Core Taxon class for more general purposes, such as identification and typification.
This decision, when implemented, means that properties with the same name in TNU and Darwin Core can have slightly different meaning. For example, tnu:genus applies to the genus part of a combination, while dwc:genus can also refer to a Taxon with a rank of genus (taxonomicStatus will have a different vocabulary as well). Also, we can rename certain properties, for example scientificName.
We spent way too much time discussing the taxonomicNameUsageLabel property. In the previous meeting it was suggested that having the property would give too much weight to the attribute, which is derived from other data, and it would be better to use something like the SKOS extended labels or the Dublin Core title attribute. I think at the meeting most people veered towards the SKOS extended label (or it least thought that this could be a use case for it). Nobody thinks we shouldn't have the attribute ("Objects need labels" (free after @ghwhitbread)). Also, different use cases might require different solutions. Therefore, we just left the property in for now.
We also had a further discussion about how the label should be formatted, especially whether the attribution for the name should be included or not. No firm conclusion on that one either.
...or Relationship Assertions.
We discussed a comment @nielsklazenga made about the hasRelationship property in the Google Doc (and in the TDWG Taxon Concept LSID Ontology) that, rather than the relationship objects, he would like to be able to have different properties for the different relationship types. It was concluded that this is a matter of different serializations.
We confirmed the properties of the Relationship class (not the name of the class, which needs further discussion).
Next meeting we will commence discussing the relationship types.
At the second meeting, all participants having been in the first meeting, we discussed various things, not necessarily all taxonomic-name-related. We agreed it would be nice to try to set up a Semantic MediaWiki site to manage the (development of the) standard and the examples; see if that works better than the current Google Doc and Sheets.