Open brandonnodnarb opened 5 years ago
@brandonnodnarb I think that this issue is an excellent candidate for the forthcoming ESIP session - https://sched.co/PtTi
Thursday, July 18 • 1:30pm - 3:00pm ESIP Geoscience Community Ontology Engineering Workshop (GCOEW)
@charlesvardeman @pbuttigieg @cmungall @brandonnodnarb does this sound reasonable?
Excellent @lewismc. I'll be there only virtually.
There are ~240 owl:sameAs relations and ~415 owl:equivalentClass relations. Obviouisly all of these aren't synonym candidates but the raw numbers without context help understand the scale of the proposal.
I think this would be an excellent candidate for the ontology engineering workshop on Thursday.
As per the discussion on #250, if skos
labels are introduced, should owl:equivalenClass
es which are obvious acronyms be deprecated with the label moving to an skos:altLabel
of the class with the full URI and label?
I have found ~80 such classes --- see this google sheet for reference.
The 'Acronyms_only' tab has the ~80 classes which appear to be obvious acronyms culled from the 'all_SWEET_owl-equivalentClass' tab. The 'all_SWEET_owl-equivalentClass' tab has all owl:equivalendClass
relationships found in SWEET listed (dump from SPARQL query).
and it looks like one goes the other way, line 22 of the all_ list has envirStandards/RfD in the right column and ReferenceDose in the left column.
There are others, see lines 73 and 74 in the Acronyms spreadsheet for example. I missed this one.
It's now been added.
As the development of SWEET initially pre-dates SKOS, there are tens, if not hundreds, of owl:equivalentClass relationships which appear to be a mechanism for adding what is essentially alternate labels (i.e.
skos:altLabel
), including synonyms and acronyms (the latter is much more prevalent).A candidate list is needed for acronyms and synonym labels.
Related to #20 in particular https://github.com/ESIPFed/sweet/issues/20#issuecomment-491018463