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

Re: owl:sameAs - Is it used in a right way?

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Thu, 21 Mar 2013 11:05:37 -0400
Message-ID: <514B21C1.8080608@openlinksw.com>
To: public-semweb-lifesci@w3.org
On 3/21/13 10:57 AM, Michel Dumontier wrote:
> Kingsley,
>  I think you raise good points. I also nominally speak of entities, 
> their attributes and the relations that hold between them. But I think 
> your diagram is somewhat misleading. URIs do denote (can stand in the 
> place of) entities of interest in order to refer to and/or describe 
> them. If you separate the identifier from what it intends to identify, 
> then you'll need another identifier for the object of interest (and 
> recurse). Hence, a information containing object (document, graph, 
> file, etc) may refer to our entity of interest by some label. the 
> document could refer to it in some statement, or may contain elaborate 
> descriptions about it.  But in no way does a document directly refer 
> to a real world object (in does so indirectly through some token - an 
> english name, identifier or uri, etc).
> m.

I don't see how my illustration conveys anything contrary to what you've 
just outlined. The parts:

1. An entity (in this case a Person) -- a real-world or non Web realm 
2. An identifier that denotes the Person -- e.g., a URI
3. A document that describes the person -- e.g. a Document comprised of 
RDF model based content in the form of an description graph (you can 
have many of these, each denoted with its own URI-URL based identifier).

Note: URI-URL is how I convey the fact that you can denote a Web 
Document using its URL. Thus, Turtle, RDF/XML, RDFa+(X)HTML etc.. 
documents (each with their own URI-URLs) can each be comprised of graphs 
describing the same entity.

> On Thu, Mar 21, 2013 at 10:27 AM, Kingsley Idehen 
> <kidehen@openlinksw.com <mailto:kidehen@openlinksw.com>> wrote:
>     On 3/20/13 10:58 PM, David Booth wrote:
>>     Thus, to be very clear, under the existing RDF Semantics
>>     specification, a given URI does **not** necessarily map to only
>>     one resource. 
>     True, but I don't think the statement above always provides the
>     clarity intended.
>     "Resource" is a synonym of "Entity" as *now* clearly stated in the
>     latest RDF concepts guide [1]. Once we get over the conflation
>     inherent in "Resource" and look towards "Entity" the issue starts
>     to get much clearer, as exemplified by RDF based Linked Data [2]
>     and its specific use of URIs to denote "Entities" while also
>     identifying their "Descriptor Documents".
>     All "Resources" aren't of the same medium. The Web, Internet are
>     mediums distinct from the medium we *refer to* as the real-world.
>     Thus, the claim that everything is a "Resource" without medium
>     specificity is one of the ultimate recipes for unproductive debate
>     and confusion, as a zillion mail threads over the years have
>     demonstrated.
>     In the context of RDF, a URI denotes an Entity.
>     In the context of RDF based Linked Data, a URI denotes an Entity
>     in a manner that enables it resolve to a Web Document (denoted by
>     its own URI which is usually a URL) that describes the denoted
>     entity (aka. URI referent) .
>     Links:
>     1. http://www.w3.org/TR/rdf11-concepts/#resources-and-statements
>     -- showcasing the critical "Resource" fix
>     2. http://twitpic.com/cbk8ul -- illustrating HTTP URI duality and
>     the Semiotic triangle .
>     -- 
>     Regards,
>     Kingsley Idehen	
>     Founder & CEO
>     OpenLink Software
>     Company Web:http://www.openlinksw.com
>     Personal Weblog:http://www.openlinksw.com/blog/~kidehen  <http://www.openlinksw.com/blog/%7Ekidehen>
>     Twitter/Identi.ca handle: @kidehen
>     Google+ Profile:https://plus.google.com/112399767740508618350/about
>     LinkedIn Profile:http://www.linkedin.com/in/kidehen
> -- 
> Michel Dumontier
> Associate Professor of Bioinformatics, Carleton University
> Chair, W3C Semantic Web for Health Care and the Life Sciences Interest 
> Group
> http://dumontierlab.com



Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen

Received on Thursday, 21 March 2013 15:06:04 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:53:01 UTC