W3C home > Mailing lists > Public > public-vocabs@w3.org > July 2013

Re: Extension of the schema.org medical vocabulary

From: Marcus Nitzschke <marcus.nitzschke@gmx.com>
Date: Mon, 22 Jul 2013 22:17:23 +0200
Message-ID: <51ED9353.1010907@gmx.com>
To: Wes Turner <wes.turner@gmail.com>
CC: public-vocabs@w3.org
Am 19.07.2013 05:55, schrieb Wes Turner:
>  > At our work group we actively use the schema.org <http://schema.org>
> medical vocabulary for
>  > modeling a health information system. However, we find that there are
> some entities that seem to be described very rudimentary. E.g. the drug
> class models the 'activeIngredient', 'administrationRoute' or
> 'dosageForm' simply as text, which doesn't provide much "semantic
> information".
>
> I assume you are referring to http://schema.org/Drug .
Right.

>
> * Are you suggesting that the range should be Text OR URL for these
> properties?
>
> * Or, are you suggesting that there should be additional,
> ontology-specific attributes for linking into more comprehensive
> pharmacodynamic ontologies?
>
I think that the activeIngredient may provide enough additional 
properties for an own class. And regarding the administrationRoute and 
dosageForm my suggestion would have been to provide some instances. But 
as the meddocs state, this is not intended by the authors. So at the end 
"Text or URL" may be an improvement nonetheless.

> There are many shared ontologies indexed at
> http://bioportal.bioontology.org/ontologies . Specifically, RxNorm may
> cover your use case:
>
> * http://bioportal.bioontology.org/ontologies/1423
> *
> http://www.nlm.nih.gov/research/umls/rxnorm/docs/2013/rxnorm_doco_full_2013-2.html
>
> It appears that the latest version of RxNorm is not yet uploaded to
> BioPortal.
>
>  > We also would like to propose a tighter coupling of Services/Business
> and Medical Entities. Currently we implement these things in our own
> ontologies which you will find here [1] bit by bit.
>  >
>  > My first question is now: Is there a broader interest/demand in
>  > improving this part of the vocabulary or are we the only ones with
> this opinion?
>
>  From http://schema.org/docs/meddocs.html :
>
>  >  Our approach is intended to be a framework for tagging known or
> novel medical concepts/entities, and optionally their relationships, as
> they appear in freeform text on the web. To manage scope, we have
> focused on markup that will help in use cases such as patients,
> physicians, and generally health-interested consumers searching for
> relevant health information. It is explicitly not our goal to replace
> existing ontology systems or to enumerate instances of medical entities,
> though our schema can link to and take advantage of existing ontologies
> and enumerations. It is also explicitly not a goal to support automated
> reasoning, medical records coding, or genomic tagging, all of which
> would require substantially more detailed (and hence high
> barrier-to-entry) modeling and markup.
>
>  > And second: Is there any ongoing collaboration at the moment towards
>  > improving the medical vocabulary or should we just start a new proposal
>  > at the wiki?
>
>  From http://schema.org/docs/meddocs.html :
>
>  > This initiative grew from a collaborative project that drew upon
> search expertise from the schema.org <http://schema.org> partners but
> also gained immeasurably through feedback from expert reviewers
> including the US NCBI; physicians at Harvard, Duke, and other
> institutions, as well as from several health Web sites. Contributions
> from the W3C Healthcare and Lifesciences group and Web Schemas community
> also helped bridge the complex worlds of Web standards, search and
> medicine/healthcare.
>
Received on Monday, 22 July 2013 20:17:52 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:28 UTC