RE: An HL7 RIM navigation language based on SPARQL?

hi peter,



interesting question and discussion.  i definitely agree with david's take
on the question that the physicians should have a nice GUI to separate them
from the details.



RIM is modeled with UML (tho it is somewhat odd that viseo is used to
visualize the models).  there is an OMG specification[1] that describes a
way to map from UML to OWL/RDF and there are always less rigorous
formulations that would do, along the lines of transforming a database
schema to RDF.



michael



Michael Miller

Software Engineer

Institute for Systems Biology



[1] http://www.omg.org/spec/ODM/1.0/



*From:* Peter.Hendler@kp.org [mailto:Peter.Hendler@kp.org]
*Sent:* Tuesday, October 02, 2012 10:54 AM
*To:* me@jerven.eu
*Cc:* eric@w3.org; helena.deus@deri.org; kerstin.l.forsberg@gmail.com;
LINMD.SIMON@mcrf.mfldclin.edu; meadch@mail.nih.gov; mscottmarshall@gmail.com;
public-semweb-lifesci@w3.org; ratnesh.sahay@deri.org
*Subject:* Re: An HL7 RIM navigation language based on SPARQL?



It's because clinicians will balk at the URIs.  The DSL would have the same
logic exaclty but all resource names and URIs would have to be replaced
with obvious business names.
Clinicians complain if they don't see exactly what they want it called.

How do I know?  I've been working on Kaiser's CMT (Convergent Medical
Terminology) system since 1995.  Our clinicians will not settle for SNOMED
preferred names, nor ICD names.  They want their own familiar terms.  We
have a local clinician interface term for everything.  In the background we
map to SNOMED.

You can get them to understand triplets, but you can't make them look at
Resource names or URIs.









*NOTICE TO RECIPIENT:*  If you are not the intended recipient of this
e-mail, you are prohibited from sharing, copying, or otherwise using or
disclosing its contents.  If you have received this e-mail in error, please
notify the sender immediately by reply e-mail and permanently delete this
e-mail and any attachments without reading, forwarding or saving them.
 Thank you.




*Jerven Bolleman <me@jerven.eu>*

10/02/2012 10:37 AM

To

Peter Hendler/CA/KAIPERM@KAIPERM

cc

meadch@mail.nih.gov, eric@w3.org, helena.deus@deri.org,
kerstin.l.forsberg@gmail.com, LINMD.SIMON@mcrf.mfldclin.edu,
mscottmarshall@gmail.com, public-semweb-lifesci@w3.org,
ratnesh.sahay@deri.org

Subject

Re: An HL7 RIM navigation language based on SPARQL?






Hi All,

Is SPARQL to difficult to teach to clinicians? I personally think its not.
What is difficult to explain is the data model (especially a HL7 compatible
one.)

Explaining a simple select once they understand triples is easy.

I love_my work = simple sentence = subject predicate object
<ch.linkedin.com/in/jervenbolleman> <
http://dictionary.reference.com/browse/love> <http://beta.sparql.uniprot.org>
= replace words by uri's

<ch.linkedin.com/in/jervenbolleman> <
http://dictionary.reference.com/browse/love> ?thingHeLoves = uri's by a
variable starting with a ?

Wrap in select

select
?thingHeLoves
where
{
      <ch.linkedin.com/in/jervenbolleman> <
http://dictionary.reference.com/browse/love> ?thingHeLoves
}
find
<http://beta.sparql.uniprot.org>

This basic concept is easily explainable in an afternoon. You will need at
least as much time to introduce any DSL as well.
The problem remains the HL7 data model. If you can explain that to anyone
in an afternoon you are my hero ;) and your DSL will need to fight that as
well. In which case it would be better to spend you time rewriting the HL7
data model into something that matches a clinicians model of his world. You
would need reasoning and/or rules to do so.

The benefit of sparql will be the capability to work with excell and or tab
delimited files that the clinician already has. Using for example bio-table
and the SPARQL 1.1. service keyword.

Regards,
Jerven

PS. I couldn't find an URI to identify my wife so had to fudge the example
;)
On Tue, Oct 2, 2012 at 7:15 PM, <Peter.Hendler@kp.org> wrote:
Mainly for Charlie and Eric but anyone who knows RIM.

There has been talk off and on for ever about a Domain Specific Language
for navigating RIM like graphs of data.  Seems to me SPARQL can already do
that.
But SPARQL is too much to teach clinicians.  So you could have a RIM
specific DSL that is like a RIMQL.  It could be nothing more than a thin
layer on top of SPARQL.

The clinician writes a RIMQL query, and it turns into SPARQL.  There's no
reason you couldn't do that with HL7 FHIR either.








*
NOTICE TO RECIPIENT:*  If you are not the intended recipient of this
e-mail, you are prohibited from sharing, copying, or otherwise using or
disclosing its contents.  If you have received this e-mail in error, please
notify the sender immediately by reply e-mail and permanently delete this
e-mail and any attachments without reading, forwarding or saving them.
 Thank you.





-- 
Jerven Bolleman*
*me@jerven.eu