DwC for OBO

[ramonawalls]

I have created OWL versions of DwC as RDF and applied for an OBO Foundry namespace for these ontologies. The goal is to have one place that anyone working on OBO ontologies can import DwC terms, so that we don't end up with may different interpretations. Reuse of these ontologies is voluntary, but doing so will help with interoperability.

OBO namespace request: OBOFoundry/OBOFoundry.github.io#1390

Please see https://github.com/BiodiversityOntologies/dwcobo for details on how the ontologies are created and what still needs to be done. If you have suggestions for improvement, concerns, or other use cases that need to be addressed please file an issue at https://github.com/BiodiversityOntologies/dwcobo/issues. The intention is for this to be a community resource.

[ramonawalls]

Some initial feedback from the OBO community was that this is somewhat outside the scope of OBO ontologies. A great suggestion was to have the OWL files of DwC as RDF officially hosted by TDWG. I would love to see this happen. In the mean time, I may create W3C IDs for them, or just have them resolve locally within the BCO repo, until TDWG makes a decision.

In case it is not clear, the ontologies I am proposing simply represent DwC as RDF in easy to import OWL files, with minimal ontological interpretation (e.g., specifying the grouping classes as domains). If TDWG wants to hose OWL files with even less ontological commitment (no domains), we could discuss that.

[baskaufs]

Just FYI, all of the standards metadata are already available as RDF dataset dumps in Turtle and XML. For example, to get the metadata about the dwc: namespace terms, http://rs.tdwg.org/dump/terms.ttl will give the RDF/Turtle, or use http://rs.tdwg.org/dump/terms with content-negotiation. For details on how the standards metadata are divided into RDF datasets, see "Retrieving machine readable metadat from the datasets". The data catalog is described in RDF at http://rs.tdwg.org/index.ttl (for Turtle) or http://rs.tdwg.org/index using content negotiation.

Not sure what you mean by "W3C IDs".

The decision was made around the time of the adoption of DwC to remove all range and domain declarations from the terms since they generated entailments that people didn't like if they disagreed with the graph model that was implied. That decision was embodied in the Standards Documentation Specification in section 4.4.2.2, which said that any properties that generated machine-computable entailments should be asserted in a separate document from the list of terms. The idea was that people could build a variety of ontologies on top of the basic bucket of terms to satisfy particular use cases. So far, there aren't any of these "extensions" that are part of any TDWG standard. Perhaps if the ABCD/DwC domain modeling exercise gets off the ground, there may be further movement on this.

[ramonawalls]

Thanks for the feedback, @baskaufs.

Regarding using the TDWG standard RDF in ontologies, it is not so simple. If you just import the turtle files to e.g. Protege, they are not interpreted correctly to use with OWL (almost everything is interpreted as an individual). You have to reformat to specify that they should be OWL data properties or OWL object properties. RDF classes are interpreted fine. @tucotuco had done the original conversion to OWL for BCO, but I wanted to script the process so it was more directly linked to the files in this repository. That is what I have done at https://github.com/BiodiversityOntologies/dwcobo. I will probably drop the request for an OBO Foundry PURL for the OWL files and resolve them some other way.

Oops. w3cID was a typo -- I meant w3id.org.

I understand about the lack of domains and ranges in DwC. That is why I did not originally propose having these ontologies in TDWG, because I figured that would cause problems. For me, not including them is a waste. I mean seriously, if the TDWG community can't agree that decimalLatitude is a property of Location, it is kind of sad, but not my fight to pick. Indeed, perhaps the ABCD/DwC modeling effort can get off the ground one day and reach some agreement on this.