Semantic Web Health Care and Life Sciences Interest Group Teleconference

07 Feb 2017

See also: IRC log


Brian_Pech, Thomas_Lukasik, David_Booth, EricP, Harold_Solbrig


Update of Linked Data Module pages -- David

dbooth: Please give feedback/suggestions!

Old module page: http://build.fhir.org/linked-data-module.html

Draft new page: http://dbooth.org/2017/ld.html

Update on Protege issues

harold: Scheduling a meeting w Stanford at the end of March to work out remaining issues.
... if you're using the import button in protege you get a box at the bottom of the screen that asks if you want help. I need to follow through with that. ("Don't be a paper clip")
... Haven't done anything on updating fhir.ttl
... Need all of the properties to be declared as ObjectProperties or DataProperties.
... If we load it currently it is OWL Full, and it cannot decide if something is an ObjectProperty.
... Treated as OWL Full it has too much in there. Original goal was to serve a purpose that shex is now serving -- types, cardinality, etc.
... Goes into OWL Full when we get into concept area, we start treating some things as both classes and individuals.
... Grahame said the goal of fhir.ttl is to support the reasoning we're doing now.
... Our use case at the moment is to take a combination of SNOMED, FHIR instance data, fhir.ttl and optionally some additional classes, and reason about FHIR individuals.
... We need things declared as ObjectProperties do to that.

dbooth: I thought we had decided to use the fhir.ttl ontology for reasoning and the shex for validation.
... I.e., not try to use the fhir.ttl for validation.

harold: I could trim down what is in the fhir.ttl for that purpose.
... Want to be sure that we keep the documentation.
... For every attribute/property there are three duplicate things: rdfs:label, dc:term, etc. Seems silly to have everything repeated 3 times.
... Looking at fhir:narrative, I have rdfs:label, dc:title, rdfs:comment and dcterms:description.
... rdfs:label needs to be narrative -- not including images.

<scribe> ACTION: Harold to update the generation of fhir.ttl to focus only on reasoning support -- not validation [recorded in http://www.w3.org/2017/02/07-hcls-minutes.html#action01]

<trackbot> Created ACTION-78 - Update the generation of fhir.ttl to focus only on reasoning support -- not validation [on Harold Solbrig - due 2017-02-14].

Documenting the owl importing sequence needed to make it work

harold: Not done yet, but working on it.

Updating the FHIR examples to include owl:Ontology and owl:imports declarations

harold: still pending. waiting on the imports sequence resolution.

What URI to use in FHIR owl:Ontology declaration


harold: At the moment, if I am loading conditionexample.ttl, it assigns the URL from which I loaded it. That seems reasonable.

dbooth: protege assigns it? harold: yes.
... Then what should we put in the FHIR examples? Anything?

harold: I need to try it more, to see if we really need it in the file. There may still be dependence on order of imports. Need to investigate more.

dbooth: Putting it in the file makes it more self-documenting.

harold: Need to investigate more.

Issue #36 Typing issues on value set references

harold: Still outstanding. Probably needs to get into the LOD profile.


harold: Currently cannot validate to this level. Might want to put it into a profile in the longer term.
... This issue is low priority, because the resource is conceptmap, which is a meta resource.
... No action at present. It will be coupled to higher priority issue of gettting typing more rigorous.

Updating Shex genjerator to handle required codings and codedelements

harold: Shex generation assumes that all values are type code, but there are several that are type coding or codedelement.
... Need to fix that in the shex generator.
... There are about 12 failing test now, because of that.

Issue 29: Concept code URI: Should we have one? What should it be?

harold: That was resolved. We're using an rdf:type arc.


Issue 16: Should fhir:reference.reference->value be a string or an RDF URI node?


harold: Resolved that one. Called it fhir:link (optional).
... And fhir:Reference.reference is the URI string

Issue 6: Downloadable FHIR ontology - Who maintains it and how?


harold: it is produced by the FHIR build process now.


<trackbot> Meeting: Semantic Web Health Care and Life Sciences Interest Group Teleconference

<trackbot> Date: 07 February 2017

Summary of Action Items

[NEW] ACTION: Harold to update the generation of fhir.ttl to focus only on reasoning support -- not validation [recorded in http://www.w3.org/2017/02/07-hcls-minutes.html#action01]

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.148 (CVS log)
$Date: 2017/02/07 22:00:57 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.148  of Date: 2016/10/11 12:55:14  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/xxx/Updating Shex genjerator to handle required codings and codedelements/
No ScribeNick specified.  Guessing ScribeNick: dbooth
Inferring Scribes: dbooth
Present: Brian_Pech Thomas_Lukasik David_Booth EricP Harold_Solbrig

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 07 Feb 2017
Guessing minutes URL: http://www.w3.org/2017/02/07-hcls-minutes.html
People with action items: harold

[End of scribe.perl diagnostic output]