search

ESIPFed / sweet

Official repository for Semantic Web for Earth and Environmental Terminology (SWEET) Ontologies
Other
115 stars 33 forks source link

Added ~5000 minerals #224

Closed brandonnodnarb closed 3 years ago

brandonnodnarb commented 3 years ago

Minerals from mindat.org as well as mineralogy textbook.

brandonnodnarb commented 3 years ago

@dr-shorthair does this help?

@lewismc would it be possible to use your wikidata code to generate likely equivalence axioms (and pull in those properties)?

nicholascar commented 3 years ago

@KellyVance @DavidCrosswellGSQ please note these large content additions to SWEET, especially in light of recent work such as https://github.com/geological-survey-of-queensland/vocabularies/pull/285

smrgeoinfo commented 3 years ago

This is in my queue to review. We've also extracted a minerals vocabulary, merging info from mindat and wikidata for the Loop3D project. Also note that the Tetherless Constellation group at RPI is working on a mineral ontology 'GEMI' in connection with their mineral evolution work

brandonnodnarb commented 3 years ago

Thanks, @smrgeoinfo. Apologies for missing this earlier. I'll get on those edits!

brandonnodnarb commented 3 years ago

Also, this

I'd suggest fixing spelling of olivine (currently is olvine, line 85)

has already happened here --- #223 but the PR has not yet met the review requirements. :)

xgmachina commented 3 years ago

I suggest using the mineral species list here as the main data source: https://rruff.info/ima/ - it is a synchronization with the IMA list here http://cnmnc.main.jp (which is only in PDF format). The mineral species list on Mindat is also following the IMA list. Another interesting topic is that Mindat includes many alternative names (group names, names no longer in use, synonyms and homonyms).

xgmachina commented 3 years ago

I sent the github link to Anirudh Prabhu - he will soon graduate from RPI and join CIW. Hopefully he can contribute to this ttl file.

smrgeoinfo commented 3 years ago

Note also the mineral ontology module I generated from the Loop3D geoscience ontology (v1. now released, but still working on generating a github tag/release and setting up URI resolution). Its mined from rruff, enhanced with information from wikidata where I could match things. Includes MindatURL, rruff ids, webmineralurl, handbookof mineralogyurl, and various other useful properties. RRUFF doesn't include uris-- the mineral identifiers are calls to a php app. I'd be quite happy to see this get merged with SWEET-- it would be pretty trivial, just change gsmin prefix to mean http://sweetontology.net/matrMineral/ and subclass from somamin:Mineral. Here's an example mineral class from the GSO :

gsmin:alabandite
  rdf:type owl:Class ;
  gsmin:chemistryelements "Mn | S"@en ;
  gsmin:crystalsystem "cubic"@en ;
  gsmin:fleischersgroup "galena"@en ;
  gsmin:handbookofmineralogyurl "http://www.handbookofmineralogy.com/pdfs/alabandite.pdf"@en ;
  gsmin:imachemistry "MnS"^^rdf:HTML ;
  gsmin:imastatus "Grandfathered|Approved"@en ;
  gsmin:iupacchemname "IUPAC: ManganeseII sulfide"@en ;
  gsmin:mindatid "min-89"@en ;
  gsmin:mindaturl "http://www.mindat.org/min-89.html"^^xsd:anyURI ;
  gsmin:rruffchemistry "Mn<sup>2+</sup>S<sup>2-</sup>"^^rdf:HTML ;
  gsmin:rruffids "R070174"@en ;
  gsmin:rruffnamehtml "Alabandite"^^rdf:HTML ;
  gsmin:rruffnameplain "Alabandite"@en ;
  gsmin:statusnotes "Called alabandina sulfurea by Don Andres Manuel del Rio as reported in: Haüy R J (1822) Manganèse sulfuré, Traité de Minéralogie 4, 268-272 Renamed alabandine: Beudant F S (1832) Alabandine, in Traité Élémentaire de Minéralogie, 2nd Edition, (Paris) 399-400"@en ;
  gsmin:structuralgroup "Rocksalt"@en ;
  gsmin:strunzcodeV10 "02.CD.10"@en ;
  gsmin:strunzlabel "Metal sulfide (M = S) with Sn, Pb, Hg, etc."@en ;
  gsmin:webmineralurl "http://www.webmineral.com/data/Alabandite.shtml"@en ;
  gsmin:wikipediadate "1822"@en ;
  rdfs:label "Alabandite"@en ;
  rdfs:subClassOf gsog:Mineral ;
.

next steps would be to link the chemistry to element URIs, and generate classes for structuralgroup, strunzcode classes, fleischer groups, and crystal systems.

dr-shorthair commented 3 years ago

Element URIs from ChEBI?

smrgeoinfo commented 3 years ago

That would be a good thing. The chebi URIs are in the GSO elements module:

gsel:aluminium
  rdf:type owl:Class ;
  gsel:atomicNumber 13 ;
  gsel:chebiid "http://purl.obolibrary.org/obo/CHEBI_28984"^^xsd:anyURI ;
  gsel:symbol "Al" ;
  gsel:wikidataid "https://www.wikidata.org/wiki/Q663" ;
  rdfs:label "aluminium"@en ;
  rdfs:seeAlso "https://www.britannica.com/science/aluminium"^^xsd:anyURI ;
  rdfs:subClassOf gsoc:Element ;
.

But we don't want to import CHEBI-- it brings a lot of baggage.

nicholascar commented 3 years ago

we don't want to import CHEBI

Well I think the solution is as per QUDT: we profile CHEBI.

Let’s make a vocabulary profile of CHEBI with just the chemicals needed and just the properties of those chemicals that we care about.

smrgeoinfo commented 3 years ago

I'm studying CHEBI to see what might reasonably be extracted for a profile. I'd follow the pattern we used for the GSO QUDT profile. This is really a separate issue, I'll move over to the Loop3D/GKM issue tracker

brandonnodnarb commented 3 years ago

Thanks, @smrgeoinfo this looks great. I had considered adding an annotation property for the formulas in an effort to be consistent with ChEBI and wikidata, but your solution works.

I'd be quite happy to see this get merged with SWEET-- it would be pretty trivial, just change gsmin prefix to mean http://sweetontology.net/matrMineral/ and subclass from somamin:Mineral.

Yes please. Happy to help if useful.

I had previously investigated Dana classification. Would that also be useful for Loop3D?

brandonnodnarb commented 3 years ago

@smrgeoinfo I (finally) incorporated the edits you suggested.

lewismc commented 3 years ago

Let's go ahead and merge this before we reattempt the skos defintions.

dr-shorthair commented 3 years ago
  1. I look forward to seeing these sorted into some kind of hierarchy
  2. Aluminum and Hematite are US spellings. Should we add British alternatives?
brandonnodnarb commented 3 years ago

@lewismc -- yeah, good call.

@dr-shorthair 1, Agreed.

  1. I think so, yes. There's also Sulfur vs Sulphur and I'm sure others. This is likely a broader issue that may need a pattern --- adding additional rdfs:label tags vs skos:altLabel vs...something else?
lewismc commented 3 years ago

Open new issue to cover these specific cases.

On Fri, May 28, 2021 at 06:54 brandonw @.***> wrote:

@lewismc https://github.com/lewismc -- yeah, good call.

@dr-shorthair https://github.com/dr-shorthair 1, Agreed.

  1. I think so, yes. There's also Sulfur vs Sulphur and I'm sure others. This is likely a broader issue that may need a pattern --- adding additional rdfs:label tags vs skos:altLabel vs...something else?

— You are receiving this because you were mentioned.

Reply to this email directly, view it on GitHub https://github.com/ESIPFed/sweet/pull/224#issuecomment-850436130, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAI4TF2OFLVYJ5XYAL7ERCLTP6OBFANCNFSM4RPTSQTQ .

--

Lewis Dr. Lewis J. McGibbney Ph.D, B.Sc Skype: lewis.john.mcgibbney