harold: Grahame fixed the RDF generation -- added a link to the metadata section. He had FHIR URI to the profile you were implementing, but forgot to put it in <>. He fixed that.
david: I'm surprised we didn't get an RDF checker in the build.
harold: I thought there was a complete RDF round tripping check in there already.
david: This exposed the lack of a QA check.
<hsolbrig> TODO: Discuss ShEx unit test w/ Graham
ISSUE: FHIR build needs RDF QA check (roundtrip?). FHIR build was generating bad RDF (though now fixed)
<trackbot> Sorry, but no Tracker is associated with this channel.
harold: I have tacit permission
to get the fixes I need for I2B2, so I can look at RDF QA when
I get to that.
... wikidata people are doing some cool things w shex right
now, looking for errors and omissions.
... Interesting side effect: When they were originally putting
it together, they were doing a standard RDF GET from the wiki,
but they weren't getting all the information.
eric: When you ask for Turtle, they made arbitrary choices about what came back, and not enough for their shex checks.
harold: I have a slurper that,
when shex needs to validate something, it gets the triples that
it needs. In conservative mode you end up with exactly the set
of triples needed for the shex expression.
... That allows you to use it on an arbitrary soup of RDF and
extract the portion that you want.
david: That's worth telling people about!
harold: Need to put together an
example to show folks. Need to host a bunch of FHIR somewhere,
to demo it.
... The power of this would be, when you have an RDF store, if
you can convert your schema into shex, you give it a subject
and it will give you the RDF that matches that model.
eric: There's a reasonable chance
we could get that functionality into wikidata.
... Andre initially wanted a shex editor, but then wrote bare
shex.
eric: UML-to-shex converter is doable.
https://rawgit.com/ericprud/XMItoRDF/master/doc/index.html
eric: drag the "m" to the text
field, and hit return.
... The "c" is the canonicalized version.
(eric demos his XML-to-RDF converter)
david: What schema is this?
eric: Schema for statistical
information
... It reads XMI, interprets datatypes and enums and
attributes, and interprets the associations.
https://ericprud.github.io/XMItoRDF/doc/ddi.shex.html
harold: There are also tools that
show UML as HTML.
... Might want to add that to the demo also.
david: we have a few use case notes from early on:
http://build.fhir.org/linked-data-module.html
See also:
http://wiki.hl7.org/index.php?title=FHIR_Ontology_Requirements#USE_CASES
http://wiki.hl7.org/index.php?title=RDF_for_Semantic_Interoperability#Use_Cases
https://www.w3.org/wiki/HCLS/ClinicalObservationsInteroperability#Use_Cases
david: the purpose of "Common Use Cases" section in the FHIR spec is to help people get started on using it. I think a use case that Harold has already been showing would be a good starting point.
harold: RDF is definitely beneficial. It gives standard URIs. Otherwise we'd have *a* mapping to I2B2, but not canonical. But with RDF we have standard URIs.
david: The work needed is to put harold's demo into tutorial form.
harold: If our paper gets accepted . . . .
david: But important to go ahead with it either way.
ADJOURNED
This is scribe.perl Revision: 1.152 of Date: 2017/02/06 11:04:15 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Succeeded: i/eric: UML-to-shex converter is doable/Topic: eric: UML-to-shex converter Succeeded: i/harold: I have a slurper that/Topic: Shex-driven RDF slurper Present: EricP David_Booth Harold_Solbrig Pawel_(briefly _muted) No ScribeNick specified. Guessing ScribeNick: dbooth Inferring Scribes: dbooth WARNING: No date found! Assuming today. (Hint: Specify the W3C IRC log URL, and the date will be determined from that.) Or specify the date like this: <dbooth> Date: 12 Sep 2002 People with action items: WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]