W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > May 2016

Re: FHIR on schema.org

From: Renato Iannella <r@iannel.la>
Date: Mon, 30 May 2016 15:43:13 +1000
Cc: "its@lists.hl7.org" <its@lists.HL7.org>, w3c semweb HCLS <public-semweb-lifesci@w3.org>
Message-Id: <FB692195-5AB3-48B5-BDD4-EC568E2C82F1@iannel.la>
To: David Booth <david@dbooth.org>

> On 28 May 2016, at 00:56, David Booth <david@dbooth.org> wrote:
> 
> It seems to me that privacy needs to be addressed at the level of protocols and policies.  What are you suggesting relevant to vocabularies, such as schema.org?

I am raising the concern that the FHIR vocabulary includes personal/private details (eg patient etc) which is *not consistent* with the scope and purpose of Schema.org

Schema.Org has no protocols/policies when it comes to the use of its vocabularies (and it does not need them, as the intent is to maximise publication of structured data).

So for privacy/policy sensitive sectors, schema.org is not the right path.

> One step toward standards convergence is to have formal semantic linkage between vocabularies.  This is essential to prevent babelization that would otherwise occur when yet another standard (such as FHIR or schema.org) is defined:
> http://xkcd.com/927/

This is ironic then? By creating the FHIR Schema.Org vocabulary, you just created the 101st standard to deal with?

> Once concepts from other vocabularies (such as FHIR) are brought into a vocabulary (such as schema.org) then the overlaps and differences between concepts become more visible, and it becomes easier for the community to reconcile 
> them and converge on a set of shared concepts.

I would argue against that. Schema.Org was never designed as a vocabulary mapping tool.
By “dumping” all the 101 health vocabularies into Schema.Org will not address mappings either.

SKOS should be used to express mappings. And should be maintained by an reputable health/clinical SDO.

Also, what use cases are trying to be solved here??

> There is a lot of visibility and institutional backing behind schema.org.  Rightly or wrongly this gives it the possibility of acting as an uber-vocabulary that spans many domains -- including healthcare -- and helps toward standards convergence.

A lot of people get captivated by the allure of Schema.Org (must be good if Google is doing it ;-)
Schema.Org is *controlled* by a small steering committee [1] (Search Engine representatives only).
Hence, it does not represent open consensus in standards - including the ability to change the schemas without notice [2].


Renato


[1] http://schema.org/docs/about.html
[2] http://schema.org/docs/terms.html
Received on Monday, 30 May 2016 05:43:45 UTC

This archive was generated by hypermail 2.3.1 : Monday, 30 May 2016 05:43:45 UTC