Minutes

VEPs"> Review of the last VEPs

MD feels the VEPs for datalink/core#counterpart and datalink/core#coderived worked quite nicely. PD emphasised how difficult it turned out to be to find well-fitting terms, and even with them having precise definitions remains important.

A VEP for Pat's interpretation of datalink#calibration

See http://mail.ivoa.net/pipermail/semantics/2020-June/002736.html

Should we just go ahead? MM volunteers to write up the VEP.

Katie would like to have more calibration-related concepts in the UAT. The longstanding issue re: calibration concepts in the UAT: https://github.com/astrothesaurus/UAT/issues/37

While this is only loosely related to what we do in datalink, it's certainly valuable to watch the UAT developments.

Registry: Who's taking the next steps with working out the proper UAT integration into VOResource?

See http://mail.ivoa.net/pipermail/registry/2020-May/005429.html

MD will start an EN for adopting the UAT. Takeup in VOResource would be in an Erratum, probably to be written between MD and TD.

Report on an effort to map current Registry subjects to the UAT

These are MD's activities with a goal to have a Semantics-based Registry Browser (SemBaReBro), which is developed at http://svn.ari.uni-heidelberg.de/svn/gavo/hdinputs/sembarebro

The start of this is to map what's already in the Registry to the UAT. Here are examples for some of the problem classes:

• Registry, OpenSkyNode, Authority: This that probably only have meaning within the VO.
• Rotational_Velocities vs. galaxy-rotation and stellar-rotation (and ~10 more like this): the existing term probably is too broad to denote a scientifically sensible concept [Ok: the term might be a valid hypernoym, the question is: "scientifically sensible"]
• Stars (and a few other hyperonyms for existing UAT terms): Terms that were controversial in the UAT but probably denote scientifically sensible concepts.
• Optical, Infrared, EUV: Wavebands used as subject keywords – I'd like to move these to coverage.
• "Observational cosmology, galaxy, cluster, evolution, infrared": We ought to contact providers who bunch together multiple terms rather than engage in probabilitstic parsing.
• Gaia: Well, HST is in the UAT... [note: activities towards a vocabulary of instruments are still underway]

What does "traverse the UAT" (from sembarebro's README) mean? It's a reference to the fact that the UAT is an organized hierarchy, so you can travel up and down the hierarchy, as well as across branches via "related concepts" [Katie]

Is this project a tool for browsing the mapping between IVOA-version and original-version for a given dictionary adopted by IVOA as "official"? MD: No – on the long the resource records should use UAT terms. On the shorter run – well, it might be that the operators of the RegTAP services agree on this and all offer rr.mapped_semantics tables. But it's far too early to even speculate whether that's a useful thing to want.

A vocabulary of messengers (for now, for VODataService waveband)

See: http://www.ivoa.net/rdf/messenger (and the VODataService in Volute rev. 5870).

[Carlo] : who will be the users of this vocabulary? MD: over and above Registry hopefully anything that will need to say what messengers were used to obtain data. And yes, solar system and theory would count.

Hence, neutrons (considering Solar System) or gravitational waves ought to go in there once there are resources wanting it.

Is it too nasty to abuse VODataService waveband for all kinds of messengers? Maybe, but I could be sold on it --Theresa

Next steps with Vocabularies 2.0?

[Carlo]: Considering the remarks from Franck on the semantic mailing list, and since I have too several remarks, I would suggest to discuss this particular point in a separate telco (in 15 days?), starting directly from this point.

That should happen before RFC. Markus will try to convene people after there's been a response to his last mail.

Gerard will try to write 1-2 paragraphs on usage of semantic vocabs in VO-DML. Happy to get help!

How do we go on with these meetings?

Markus would like to have this kind of thing every two months -- which people can probably stand. We'll have a focused Theory-Semantics meeting sooner, though.