A more sophisticated ontology mapping

[lukecampbell]

This has been discussed in multiple issues. It would be nice to see, as we harvest we map the variable metadata to a well defined ontology (MMI). This would help us with searching, mapping, contexts etc.

[emiliom]

I'm intrigued to hear more about this. Do you mean a mapping of an ontology data structure within the Catalog code (including linking and harvesting of vocabulary and vocabulary mappings from MMI), or more extensive ontological mappings among vocabularies on MMI?

For reference, I'd highlight issue #72 (including the link to the Use Cases for Vocabulary Search Capability document from @robragsdale and Sara Haines) and issue #322. See also ioos/system-test#128 and ioos/system-test#129.

Also in case it's of use, after a bit of brainstorming between MMI's @graybeal and I early this year, I started compiling some MMI-relevant Python code on a little repo of mine; there's nothing in here that's fundamentally new and hasn't been posted by Sarah or others. Just a compilation of some IPython notebooks, and a small list of generic but relevant semantic needs inspired by IOOS and other contexts. I haven't been able to find the time or motivation to flesh it out in several months. With the right motivation and contexts, I could put some effort into it again.

I'm interested in having some discussions around these topics at the upcoming IOOS DMAC meeting. Maybe even a bit of coding, if the stars align. @lukecampbell, are you coming to the meeting?