W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > January 2013

Re: Opening Walled Gardens: RDF / Linked Data as the Universal Exchange Language of Healthcare

From: Frank Manola <fmanola@acm.org>
Date: Tue, 15 Jan 2013 10:32:34 -0500
Message-id: <9492BEEC-6689-47CC-A88A-939D86C0B280@acm.org>
Cc: RebholzSchuhmann <d.rebholz.schuhmann@gmail.com>, Joanne Luciano <jluciano@gmail.com>, public-semweb-lifesci <public-semweb-lifesci@w3.org>, Michel Dumontier <michel.dumontier@gmail.com>, Conor Dowling <conor-dowling@caregraf.com>, Rafael Richards <rmrich5@gmail.com>
To: David Booth <david@dbooth.org>
David--

A few quick comments below.

On Jan 15, 2013, at 9:34 AM, David Booth wrote:

> Hi, and thanks for your comments!
> 
> On Tue, 2013-01-15 at 12:58 +0000, RebholzSchuhmann wrote:
>> Hi,
>> 
>> don't know how someone reads this, who does not know all these benefits 
>> anyways. Reads as if you are selling RDF to somebody who knows half-way 
>> the benefits of RDF.
> 
> Yes, we decided that we simply didn't have time to write a long document
> that more fully explained the benefits.  (And actually, I'm not sure
> that would have been effective anyway, as people usually need to see new
> ideas multiple times before they understand them.)  We don't expect this
> one comment to change anyone's mind, but our hope is that it will get
> people to start looking in this direction.

The shortness of this is appropriate under the circumstances, but I'm sure you know that the points you make are somewhat more "nuanced" than they are made out to be here.

> 
>> It would have made sense to be more precise on the privacy and security 
>> issues. 
> 
> Good point, since that's such a fear-inducing topic. 
> 
>> Neither RDF nor XML have been developed to address privacy / 
>> security, and either one is highly important in healthcare systems.  Do 
>> you have even stronger arguments for privacy and security issues?
> 
> Not really.  The main argument is that the same techniques that are
> currently being used can still be used.  Privacy and security issues are
> orthogonal to information representation choices, or at least they
> should be.  One could make arguments about the potential for using RDF
> to reason about access permissions, but I think that would be somewhat
> specious, because RDF could be used for that purpose even if the
> information representation language were not RDF.
> 
> Perhaps one security argument we could make is that RDF reduces
> complexity, by providing a uniform information representation language,
> and as we all know, complexity reduces security because it increases
> vulnerabilities.  But I don't think that's a particularly strong
> argument either.
> 
> Do you have any ideas about privacy and security with respect to RDF?

It seems to me that you could be more "precise" on the privacy and security issue by not introducing it with the bare assertion that "RDF is secure".  This has an easy counter-argument (as you note, it's orthogonal to what RDF is about), and thus weakens the presentation.  What you say in explanation of that bullet is fine.  Perhaps the point should be simply along the lines of "RDF works with existing privacy and security mechanisms".  

On another subject, it seems to me you could avoid making such a sharp contrast between using RDF and using XML.  There are lots of people who are invested in XML, not inappropriately, and RDF is perfectly compatible with using XML (and an XML representation for RDF is part of the "standard", although that's not the only way to represent RDF in XML), just as using RDF doesn't mean giving up relational databases (as you note). It would take longer to explain the why this is true but, as noted earlier, that's true of other points you make in your response too.  

--Frank
Received on Tuesday, 15 January 2013 15:33:05 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:01:17 GMT