See also: IRC log
tony: Datatype quantity is modeled differently than CodeableConcept. It's like a CodeableConcept but flattened.
eric: There is something that looks similar to CodeableConcept, but isn't. Right?
tony: Yes.
... Unit of measure should used the same structure, but it
didn't.
dbooth: Should we give feedback on that?
tony: Yes. I asked grahame but haven't heard back.
eric: Do these have a min and max of 1 coding?
tony: I think so, like a restriction on CodeableConcept, but it has text also. We could model it as a restriction on CodeableConcept.
<scribe> ACTION: Tony to document the issue about dataype quantity being modeled differently from CodeableConcept, and check with Lloyd [recorded in http://www.w3.org/2016/01/19-hcls-minutes.html#action01]
<trackbot> Error finding 'Tony'. You can review and register nicknames at <http://www.w3.org/2014/HCLS/track/users>.
dbooth: In talking with Grahame, he mentioned that FHIR has some indexes, and they start with 0. This was news to me. For FHIR RDF we had decided to start the index at 1, but for consistency sake, I think we should revisit that question.
eric: +1
tony: I found one problem with
fhir:index -- I used it in two ways. I used it for schema
sequence, and that was a problem. Same name for an annotation
and an object property causes a property.
... I will rename it to something else.
eric: In theory it should not be a problem. It is OWL punning, but there may be bugs that are related to it.
tony: It's not the punning, it's the loading process when using Turtle. The loader at load time doesn't now what it is.
eric: There's a set of
recognition rules for interpreting OWL serialized in
Turtle.
... But ultimately, punning only causes trouble.
tony: It was only the loading that was the problem.
dbooth: We'll re-decide the index (0 or 1) next time we have a quorum.
<scribe> ACTION: dbooth to ask grahame for index 0 FHIR examples [recorded in http://www.w3.org/2016/01/19-hcls-minutes.html#action02]
<trackbot> Created ACTION-42 - Ask grahame for index 0 fhir examples [on David Booth - due 2016-01-26].
dbooth: We should be using https://github.com/w3c/hcls-fhir-rdf/ repository now.
dbooth: Need to get it into xhtml and into github
tony: Can't find documentation on
what the xhtml should look like.
... Want it to be compatible with the FHIR build process
http://wiki.hl7.org/index.php?title=FHIR_Build_Process
eric: If you are doing pull requests against xhtml then you should use a tool that does not reformat it, so that github does not report spurious changes.
<scribe> ACTION: Tony to send side-by-side v12 to Eric for conversion to xhtml and pushing to github [recorded in http://www.w3.org/2016/01/19-hcls-minutes.html#action03]
<trackbot> Error finding 'Tony'. You can review and register nicknames at <http://www.w3.org/2014/HCLS/track/users>.
Tony's side-by-side document: http://wiki.hl7.org/index.php?title=File:FHIR_RDF_Sample_side_by_side_comparisons.pdf
line 76
tony: It's a flat definition of
elements, in <snapshot>
... line 122, lloyd says there may be an overarching
cardinality constraint that would further limit this.
... Reference.display will be come an object property (line
157), whereas Reference becomes a class.
... line 168 cardinality will translate into owl cardinality on
line 206
eric: This is the same thing we
did with ShEx
... If we have some code that walks through the structure def
and spits out OWL, we'd want to document the mapping and test
it.
tony: yes. In this doc (line 222)
we show an example.
... Line 230 (sec 1.3.6) is the meat of it.
eric: Do we have slicing
requirements in FHIR itself, or only in profiles?
... Even if we cannot write slicing requirements in OWL, we can
still write them using our own namespace, and potentially
handle them in SPARQL.
tony: Sec 1.3.6 is the meat of
what we need to do.
... <path ..> will sometimes be a class, sometimes a
property.
dbooth: If Structure Definitions can be sent as FHIR messages, then they need to be round-trippable.
tony: No they don't.
eric: The ROI for making profiles interoperable in RDF is much less than for making patient instance data interoperable.
dbooth: Concerned that that would
make FHIR RDF a second class serialization.
... Not on par with FHIR XML or FHIR JSON.
tony: If RDF and OWL could be an authoring for profiles, then you've opened a new available tooling resources.
dbooth: For transmission
purposes, it might be good enough to send the structure def as
one big string.
... I don't want anyone to disparage RDF for not being able to
represent all of FHIR.
eric: XML Schema + Schematron would probably handle the expressivity.
harold: their metamodel is
self-defined
... Therefore we could transmit it as an ABox
ADJOURNED
eric, re: ACTION: Dale to introduce EricP to Aegis folks about testing the FHIR XML<->RDF translation
eric, anything new on that?
<ericP> http://hl7-fhir.github.io/observation.html
http://wiki.hl7.org/index.php?title=File:FHIR_RDF_Sample_side_by_side_comparisons.pdf
sec 1.3.6
tony: <strength> can be: required | extensible | preferred | example
<ericP> http://www.w3.org/2013/12/FDA-TA/tests/RenalTransplantation/
dbooth: I don't really like the
owl:Ontology in everything. It seems like its sole purpose is
to make protege happy. But I guess it's harmless to keep
it.
... How do we generate the URI for line 275 and 273, in order
to be fully round trippable?
<scribe> ACTION: Tony to see if owl:Ontology can be declared on a blank node [recorded in http://www.w3.org/2016/01/19-hcls-minutes.html#action04]
<trackbot> Error finding 'Tony'. You can review and register nicknames at <http://www.w3.org/2014/HCLS/track/users>.
eric: the URIs on lines 273 and 275 should be different.
dbooth: in theory, lines 273 and
275 are URI squatting
... Suppose you are translating a FHIR XML file to FHIR RDF.
What URI would be used for lines 273 and 275?
harold: We shouldn't be deriving that URI from outside information. It should be inside the FHIR document itself.
ISSUE: How to determine the base URI for a FHIR document, and is it round trippable?
<trackbot> Created ISSUE-18 - How to determine the base uri for a fhir document, and is it round trippable?. Please complete additional details at <http://www.w3.org/2014/HCLS/track/issues/18/edit>.
ADJOURNED
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) No ScribeNick specified. Guessing ScribeNick: dbooth Inferring Scribes: dbooth Present: Thomas_Lukasik Tony_Mallia Harold_Solbrig David_Booth EricP Found Date: 19 Jan 2016 Guessing minutes URL: http://www.w3.org/2016/01/19-hcls-minutes.html People with action items: dbooth tony[End of scribe.perl diagnostic output]