- From: Dan Brickley <danbri@danbri.org>
- Date: Tue, 22 May 2012 16:34:13 +0300
- To: Aaron Brown <abbrown@google.com>
- Cc: Renato Iannella <ri@semanticidentity.com>, "Lin MD, Simon" <LINMD.SIMON@mcrf.mfldclin.edu>, Matthias Samwald <matthias.samwald@meduniwien.ac.at>, "public-semweb-lifesci@w3.org" <public-semweb-lifesci@w3.org>
On 22 May 2012 16:20, Aaron Brown <abbrown@google.com> wrote: > On Tue May 22 00:15:04 GMT-400 2012, Renato Iannella > <ri@semanticidentity.com> wrote: >> >> >> On 22 May 2012, at 10:05, Aaron Brown wrote: >> >> The schema wasn't designed with that use case in mind -- it's focused on >> marking up public information on the web for search use cases, not for >> coding or exchange of clinical data. It might be extensible to a version of >> that use case, if complemented with requirements for use of well-defined >> coding systems tied to external enumerations (like ICD or Snomed, RxNorm, >> etc). But it'd probably be better to use a purpose-built representation for >> clinical data transfer, like HL7 CDA / CCD or CCR, coupled with a >> patient-readable form of the discharge notes; standards like these offer >> greater precision in the specification of clinical data. >> >> >> It would be useful to add this design constraint to the introduction (at >> the moment it says "The scope of this schema is broad…") > > > Sure. That comment was intended to be about the scope of entities covered, > not the scope of use, as I had assumed that folks would be looking at this > in the context of schema.org, which is about web markup. I'll clarify in my > next set of updates. > >> >> And perhaps define MedicalEntity as disjoint from Person ;-) > > > Not sure I understand where the confusion is here...Person is a separate > type under Thing, so they should be disjoint by definition. We don't say anything like that at a schema.org level. Sometimes classes have common members, sometimes they don't. If you want us to record inter-class disjointness relationships within schema.org we can look into that, but for now don't assume it "comes out of the box" with different Thing subclasses being automatically considered disjoint. cheers, Dan
Received on Tuesday, 22 May 2012 13:35:07 UTC