[Proposal] Replace samplingType and observationType with sdo:additionalType

[ashleysommer]

My understanding is samplingType andobservationType were included in the TERN ontology to allow a kind of soft-typing mechanism, using more specific types from a vocabulary rather than adding new SubClasses to the TERN ontology.

There popular and increasing use of sdo:additionalType from schema.org that does the same thing, it is becoming the de-facto standard way of soft-typing across many data models.

The predicates samplingType and observationType in TERN ontology are not widely used, I don't know of any real world published datasets that use them. Most simply use/(abuse?) usedProcedure to classify different kinds of Samplings and Observations.

I propose to remove samplingType and observationType from TERN ontology and replace them with sdo:additionalType.

Aside, this would be a good opportunity to remove dcterms:type profiling from the TERN ontology if it was an implementation detail and wasn't supposed to be in the published ontology.

[ashleysommer]

There remains a question about the use of sdo:additionalType on a tern:Sample.
Sample is a kind of a FeatureOfInterest, and TERN ontology uses tern:featureType to soft-type Features.
The correct thing to do here would be also replace featureType with sdo:additionalType, however tern:featureType is extensively used, and is hard to move away from.

[smguru]

Yes can be removed, I don't think we are using both properties. FYi @javiersan18 @junrongYu