20 May 2014

See also: IRC log


EricP, Neda, Ingeborg, Tony, Mehmet, David_Booth
DBooth (by default)


ericp, are you here?

mehmet: want to hear more about RDF for Semantic Interop

ingeborg: Investigating OMOP (Observational Medical Outcomes)

Ontology developed by Chimezie Ogbuji

Tony: It's POMR (Problem oriented medical records)
... Interested in looking at an ont that represents FHIR. not necessarily the transmissions of FHIR payloads, but underpinnings and where it might migrate. i was looking differently from what Claude and Josh were looking at.
... Discussed this at HL7 meeting last week.

mehmet: That ontology might be still used in Cleveland Clinic.

Investigating OMOP

ingeborg: similar to the Sentinal program, but uses a different data model.
... Would sem web be used to underly and link concepts?
... Interested in Medra / snomed mapping. Looks like OMOP may do that.


tony: i downloaded the snomed vocab in RDF. it's huge: 100MB.
... failed to load in protege -- too big.

ingeborg: they load the entire snomed. Using oracle create table language. constrain it to clinical section of snomed.

<scribe> ACTION: Ingeborg to investigage OMOP and report back [recorded in http://www.w3.org/2014/05/20-hcls-minutes.html#action01]

RDF for Semantic Interop

david: who might be interested in RDF for Sem Interop
... ?

tony: Peter Hendler is working with RDFox triple store
... From Oxford Univ
... peter criticized the use of the RIM.

eric: RIM is for expressing any problem.
... Act relationships are the analogy for how you would express causation
... They invented a graph, and it had Act relationships, like arcs in RDF, and specialized arcs for roles and participations.
... Captures who did what to whom, but also captures causation. TermInfo page talks about how you represent a particular diagnosis with causation and severity -- how to represent it in SNOMED and RIM.
... Type codes are fairly well debugged, and they're exactly what you're looking for.

Tony: This leads to ORIM

Eric: I worked with Lloyd on the ORIM, and it's just a MIF expressed as RDF.
... There's the concept of a relatinoship which would be an arc as RDF.
... Action that's an observation, another that says behavior X was caused by Y. If you skip the date and other things you can write it as an arc.

tony: Is the relationship modeled as an entity? An object, as opposed to a property?

eric: Yes, it's an entity with a source, target and a type code. What we'd like is just that type code.

tony: Where is the ORIM?

eric: I'll dig up a copy from my harddrive.
... It's on HL7 forge site.
... Lloyd as done all of the work on it to date.
... Want to talk about FDA in two weeks.

david: I should be able to talk more about the RDF for Sem Interop draft charter next week. I plan to update it a bit, per discussions so far.


<scribe> scribe: dbooth

Summary of Action Items

[NEW] ACTION: Ingeborg to investigage OMOP and report back [recorded in http://www.w3.org/2014/05/20-hcls-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014-05-20 15:35:37 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found Scribe: dbooth
Inferring ScribeNick: dbooth
Default Present: DBooth, +1.202.528.aaaa, +1.469.226.aabb, Mehmet, Ingeborg, Neda, +1.978.794.aacc, Tony, EricP
Present: EricP Neda Ingeborg Tony Mehmet David_Booth
Got date from IRC log name: 20 May 2014
Guessing minutes URL: http://www.w3.org/2014/05/20-hcls-minutes.html
People with action items: ingeborg

[End of scribe.perl diagnostic output]