OR13
commented
3 years ago We should declare periodic table elements in a traceability vocabulary. DARPA did this work in 2003:
I assume you mean don't?
Closed msporny closed 3 years ago
OR13
commented
3 years ago We should declare periodic table elements in a traceability vocabulary. DARPA did this work in 2003:
I assume you mean don't?
msporny
commented
3 years ago I assume you mean don't?
Doh, yes, I meant "we should not declare periodic table elements in a traceability vocabulary."
I just changed the original issue text to fix that.
OR13
commented
3 years ago @msporny I would prefer to define them, even if its just a pass through definition as we have done for PostalAddress....
https://transmute-industries.github.io/traceability-vocab/#postal-address-0 -> https://www.gs1.org/voc/PostalAddress -> https://schema.org/PostalAddress
The problem is I have not found a single example that is helpful to "pass through too"... I am, and forever will be... against term definitions that do not resolve (schema.org is 500 right now.... lol)
msporny
commented
3 years ago The problem is I have not found a single example that is helpful to "pass through too"...
Yes, that tends to be a consistent problem with these "scientist-created" vocabularies... extremely complex because they need the complexity. :)
I am, and forever will be... against term definitions that do not resolve (schema.org is 500 right not.... lol)
What!? You must be mistaken... Google is never down. stares at Error: Server Error :P
On a somewhat related note, GS1's vocabulary documents look really good...
https://www.gs1.org/voc/PostalAddress
We should steal the layout... perhaps move away from ReSpec (which was always meant to be a temporary solution). /cc @philarcher
philarcher
commented
3 years ago The problem is I have not found a single example that is helpful to "pass through too"...
Yes, that tends to be a consistent problem with these "scientist-created" vocabularies... extremely complex because they need the complexity. :)
I am, and forever will be... against term definitions that do not resolve (schema.org is 500 right not.... lol)
What!? You must be mistaken... Google is never down. stares at Error: Server Error :P
On a somewhat related note, GS1's vocabulary documents look really good...
https://www.gs1.org/voc/PostalAddress
We should steal the layout... perhaps move away from ReSpec (which was always meant to be a temporary solution). /cc @philarcher
I wish I could claim credit for the GS1 Web Voc layout, but that accolade goes to https://github.com/orgs/gs1/people/mgh128. I know it's driven by PHP which is why we can't use it directly in the Web Voc's own GH repo at https://github.com/gs1/WebVoc
msporny
commented
3 years ago /me Github crawls https://github.com/mgh128/ ends up at another helpful thing for this vocab -- https://github.com/datasets/unece-units-of-measure
msporny
commented
3 years ago While we have your attention @philarcher -- From an RDF vocabulary perspective, what does GS1 do for Units of Measure?
OR13
commented
3 years ago @mgh128 thanks!
OR13
commented
3 years ago 
mgh128
commented
3 years ago @msporny Hi Manu - GS1 usually uses UN ECE Rec 20 units of measure throughout its standards, even though they're not yet very Linked Data friendly. While QUDT 2.0 was taking a while to appear and because we're adding sensor data to our GS1 EPCIS 2.0 standard for supply chain traceability data, we've prepared this unit converter for the Rec 20 unit of measure codes ( https://github.com/gs1/UnitConverterUNECERec20 ) to deal with the situation where one organisation might record some sensor data using one unit of measure (e.g. 'FAH' for degrees Fahrenheit) and another organisation sends a query asking for event data where the temperature threshold exceeded a value expressed in a different but interconvertible unit (e.g. 'CEL' for degrees Celsius). Online demo page at https://gs1.github.io/UnitConverterUNECERec20/ If you have any questions or suggestions on that, please let me know
Within the GS1 Web vocabulary at https://www.gs1.org/voc the property https://www.gs1.org/voc/unitCode within the class https://www.gs1.org/voc/QuantitativeValue expects a UN ECE Rec 20 unit code string.
mgh128
commented
3 years ago I wish I could claim credit for the GS1 Web Voc layout, but that accolade goes to https://github.com/orgs/gs1/people/mgh128. I know it's driven by PHP which is why we can't use it directly in the Web Voc's own GH repo at https://github.com/gs1/WebVoc
I've started preparing an updated of the online documentation tool for the GS1 Web vocabulary. The new version is all client-side written in Vue.js, still data-driven by the JSON-LD dataset for the GS1 Web vocabulary. Unfortunately a few other things are higher up my to-do list for GS1 - but when it's done, the plan is that we'll put the new version of the tool as a GitHub.io page for https://github.com/gs1/WebVoc - and at that point, you're welcome to 'steal' / modify a copy for other purposes. I'll try to write it in a way that it's reusable. The current tool at https://www.gs1.org/voc is a rather unwieldy mix of server-side PHP code and client-side JavaScript code from a few years ago, to ensure that it worked with / without JavaScript enabled in the browser or search engine crawler.
msporny
commented
3 years ago The new version is all client-side written in Vue.js, still data-driven by the JSON-LD dataset for the GS1 Web vocabulary. ...I'll try to write it in a way that it's reusable.
Brilliant! This would be perfect and is aligned with a development environment that much of us are familiar with... if we time this right, we could get multiple folks hacking on the code at the same time. There is a cohort of funded companies in the US that would need that sort of tool in the next 6-9 months. If we do this right, we should be able to generalize it for all Linked Data vocabulary publishing at W3C (and GS1?!). In any case, getting ahead of ourselves... looking forward to anything you can share sooner rather than later... we'd be interested in contributing to the code base (as we're going to end up writing something very similar out of necessity).
Well done on GS1 SmartSearch vocabulary btw... it's a shining example of a Linked Data vocabulary. Very much a fan. :)
OR13
commented
3 years ago amazing how hard it is to find a stable defintion for "atomic Number" or "atomic symbol".... I have spent hours on https://lov.linkeddata.es/dataset/lov
so far,
are the best things I have found.
msporny
commented
3 years ago are the best things I have found.
@kidehen and @TallTed -- how stable is dbpedia.org? What's the governance model around it? If we started using terms in it for international standards, what do you expect the arguments against it would be?
OR13
commented
3 years ago seems like chebi, may actually work...
of course all the URLs don't go anyhere, and elements are identified via CHEBI IDs....
https://www.ebi.ac.uk/chebi/searchId.do?chebiId=CHEBI:155866
mgh128
commented
3 years ago I found this:
http://qudt.org/vocab/quantitykind/AtomicNumber
On Sat, 3 Oct 2020 at 16:19 Orie Steele notifications@github.com wrote:
amazing how hard it is to find a stable defintion for "atomic Number" or "atomic symbol".... I have spent hours on https://lov.linkeddata.es/dataset/lov
so far,
are the best things I have found.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/transmute-industries/traceability-vocab/issues/1#issuecomment-703119447, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABSXRLY6YGXLIE22Y6SIJN3SI46IDANCNFSM4SAOV4PQ .
OR13
commented
3 years ago thanks!
I am considering generalizing beyond "elements" to molecular structures... https://blogs.ch.cam.ac.uk/pmr/2009/04/05/cas-and-inchi-who-can-assign-identifiers/#:~:text=CAS%20states%20its%20numbers%20refer,as%20more%20than%20one%20substance.
OR13
commented
3 years ago also useful... https://pubchem.ncbi.nlm.nih.gov/
OR13
commented
3 years ago msporny
commented
3 years ago Hey @HolgerKnublauch, how stable do you think the qudt.org site is? If we used it as the basis for a global standard vocabulary, would you have any misgivings? Could we expect anyone to object based on how the site is operated from a governance perspective?
mgh128
commented
3 years ago Within GS1 EPCIS 2.0 we're referencing InChi key values to identify chemical substances in sensor data.
More about InChi key at : https://en.m.wikipedia.org/wiki/International_Chemical_Identifier
On Sat, 3 Oct 2020, 17:45 Orie Steele, notifications@github.com wrote:
thanks!
I am considering generalizing beyond "elements" to molecular structures... https://blogs.ch.cam.ac.uk/pmr/2009/04/05/cas-and-inchi-who-can-assign-identifiers/#:~:text=CAS%20states%20its%20numbers%20refer,as%20more%20than%20one%20substance .
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/transmute-industries/traceability-vocab/issues/1#issuecomment-703131278, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABSXRL6X34D742KPYPA7W6DSI5ILLANCNFSM4SAOV4PQ .
msporny
commented
3 years ago Within GS1 EPCIS 2.0 we're referencing InChi key values to identify chemical substances in sensor data.
Could you give an example of this when doing something like an rdf:type? My guess is:
<_bn:1> <inchi:element> "H" .and I was hoping for:
<_bn:1> <rdf:type> <inchi:H> .@msporny
https://pubchem.ncbi.nlm.nih.gov/#query=InChI%3D1S%2FTi
<chebi:formula rdf:datatype="http://www.w3.org/2001/XMLSchema#string">Ti</chebi:formula>
<chebi:inchi rdf:datatype="http://www.w3.org/2001/XMLSchema#string">InChI=1S/Ti</chebi:inchi>
<chebi:inchikey rdf:datatype="http://www.w3.org/2001/XMLSchema#string">RTAQQCXQSZGOHL-UHFFFAOYSA-N</chebi:inchikey>see also: http://registry2.it.csiro.au/def/op?_format=jsonld&_view=with_metadata
OR13
commented
3 years ago msporny
commented
3 years ago Hrm... all strings... no rdf:types in sight for things on the periodic table. I'm not a chemist by any stretch of the imagination, so I'm probably missing something obvious. chebi/inchi/qudt do seem like the leading contenders for the use case.
In any case, my goal with this issue was to avoid work done elsewhere and I think @OR13 is on the trail so will disengage at this point leaving him to brave this particular rabbit hole with the rest cc'd on this thread... hope it isn't too deep. :)
OR13
commented
3 years ago no plz halp!
lol, i think this is good enough for now: https://transmute-industries.github.io/traceability-vocab/#substance
schema.org + chebi => any substance that can be expressed with inchikey
mgh128
commented
3 years ago At https://github.com/gs1/EPCIS/blob/master/JSON/WithSensorData/SensorDataExample8.json you can see one of our draft examples using InChi Key (see line 91).
However, we probably should change the value to instead use something like https://identifiers.org/inchikey:RYYVLZVUVIJVGH-UHFFFAOYSA-N
so that instead of us writing "chemicalSubstance":" urn:epcglobal:cbv:inchikey:CZMRCDWAGMRECN-UGDNZRGBSA-N"
we'd write "chemicalSubstance" : " https://identifiers.org/inchikey:CZMRCDWAGMRECN-UGDNZRGBSA-N"
I'll recommend that we make that change to our drafts - will mention it in our GS1 EPCIS call next week - and by e-mail to the chapter editors / core team before then.
( We are trying to move away from URNs (URIs that Resolve Nowhere) and drag our GS1 standards into the 21st century - but it's taking time to win everyone over! )
On Sat, Oct 3, 2020 at 6:04 PM Manu Sporny notifications@github.com wrote:
HolgerKnublauch
We should not declare periodic table elements in a traceability vocabulary. DARPA did this work in 2003:
http://explore.dublincore.net/learning_resource/periodic-table-in-owl/
There is also the Ontology of Chemical Elements, which is current and put together by biometical ontologists:
https://bioportal.bioontology.org/ontologies/OCE/?p=summary
At the very least, we really should involve chemists, physicists, etc. to determine how this work should be done... but not to the degree that this turns into a multi-year effort. Taking the two ontologies above and doing a simpler one that merges the common terms expressed in each might work as well -- but could be a significant undertaking on its own.