I've been working on a new version of a Data Type model for consideration by the working group. We're testing in the context of an EarthCube Building Block project (Digital Crust, the registry is called 'Information Exchange Registry' in the architecture diagrams) for use in registering heterogeneous tabular datasets to enable (in the goal state) more automated data integration.
A test instance was deployed at (now offline)[http://23.23.228.241:32781/class/_data-object ] (note that attribute element bindings have only been implemented for a few data objects, e.g. 'cPHydro observation data' or 'Geothermal Power Plant Occurrence').
The ontology is registered in a test ESIP ontology registry at http://cor.esipfed.org/ont/?uri=http://resources.usgin.org/uri-gin/dtr/o...
Maintenance is in GitHub at https://github.com/usgin/digital-crust-LDR.
I'm looking forward to discussions in Denver.
cheers
steve
Author: Matthew Jones
Date: 26 Jul, 2016
Thanks, Steve. Interesting work, and thanks for letting us know.
In looking over the ontology, I notice a lot of conceptual overlap with
some existing ontologies, in particular with OBOE (
https://github.com/NCEAS/oboe), OML (O& M Lite:
http://www.semantic-web-journal.net/system/files/swj890.pdf), and BCO (
http://dx.doi.org/10.1371/journal.pone.0089606). Have you done a
comparison with these existing observation ontologies, and have you
identified where your goals/needs may differ? A group of us (me, Simon
Cox, Ramona Walls, Rob Guralnick, Mark Schildhauer, Pier Luigi Buttigieg,
Adam Shepherd, and Bryce Mecum) have been working on an alignment among
these existing ontologies trying to show where they are isomorphic and
where they introduce truly new concepts (
https://github.com/Semantic-Observations/obs-models/blob/master/ontologi...)
-- this is very much a work-in-progress alignment, but the (super messy)
overview figure (
https://github.com/Semantic-Observations/obs-models/blob/master/figures/...)
should give you a decent idea of the linkages we are seeing among models.
It would be interesting to know where your new data type model fits into
this existing landscape (I see several potentially corresponding classes,
but would need to study your model more deeply to really grok it).
Looking forward to the discussion at RDA.
Matt
On Tue, Jul 26, 2016 at 9:55 AM, srichardUSGIN <
***@***.***> wrote: