Re: [ontolog-forum] Requesting Opinions on the Benefits of Predicates as

[Kingsley Idehen <kidehen@xxxxxxxxxxxxxx>]

On 6/11/14 6:56 PM, Kingsley Idehen wrote:

On 6/11/14 2:16 PM, Barkmeyer, Edward J wrote:

Kingsley,

 

you wrote:

If each part of a statement (e.g., an RDF triple) is denoted by a derefencable URI (e.g., an HTTP URI) you end up with a much deeper, web-like,  and logically comprehensible graph, to both humans and machines. Generally, this concept is what's referred to as RDF [1] based Linked Data [2] and the result is a Semantic Web [3].

Well, you get a ‘semantic network’, if that is what you mean by “Semantic Web”. 

Yes.

Basically, from my diagram collection covering computer [1], document [2], data [3], and semantic data (or Semantic Web) networks [4] .

The problem is that the edges of the graph you get are all labeled ‘also appears in’ or ‘more about’, which does not convey much in the line of semantics.

This is an issue, and some of my working diagrams should make this issue quite visible [4][5] .

  The ancient and venerable library science equivalent is “Keyword in context”.  You have to interpret the triples (reference in position relative to other references) to obtain the semantic value.

Yes.

If I understand the model you propose, the “nodes” of your graph are the triples, not the IRIs.

Not proposing that since reification already exists in RDF albeit rarely used, following lots of usual confusion etc..

When RDF statements are reified and presented alongside the others, something I will eventually illustrate, since, my diagrams take you to the point of "implied denotation" meaning: you don't seen any explicit denotation relations in my Linked Data based Glossary document, and the statements themselves aren't reified or treated as first-class entities.

I do have some live examples of reification [6][7], but (as already stated) this aspect of RDF needs to be incorporated into my glossary document as an expansion on the current Semantic Network/Web illustration.

  One of the issues in Linked Open Data is whether the “link” is the semantics-free link between two occurrences of the same IRI, or it is the ‘verb’ in the triple that connects the IRIs. 

In TURTLE:

<#LinkedDataURI>
<#implicitlyDenotes> [ a owl:Thing;
                                    <{some-relation-denoted-by-http-uri}> <{some-relation-object-denoted-by-http-uri}> ..
                                    ] .

Both camps have apparently found it useful to co-opt the LOD term.

Which is why I try to use the qualification RDF based Linked Data :-)

 

As Mark Linehan pointed out, you can treat any predicate as a class of states.  You need additional predicates to describe roles in the states, but you can have as many roles and associations as you like.  There is a lot to be said for

  A Relation B,

and different lot to be said for

  Relation R1, A SubjectOf R1 (aka R1 Agent A), B ObjectOf R1 (aka R1 Patient B)

But the one thing that can be said is that using Relation in BOTH ways will almost certainly confuse the semantics of the graph. 

Yes, if denotation is implicit (the commitment RDF processors make, at the very least) and reification isn't part of the narrative.

Links:

[1] http://bit.ly/1cjYwqN -- DNS based Linked Computer Network (Internet)
[2] http://bit.ly/1e0pwdI -- HTTP based Linked Document Network (World Wide Web)
[3] http://bit.ly/INv6ag -- Linked Open Data Network (LOD Cloud)
[4] http://bit.ly/1lHD31h -- initial simplified Semantic Linked Open Data Network (Semantic Web, modulo denotation )
[5] http://bit.ly/1qRRYcn -- revised Semantic Linked Open Data Network (Semantic Web, with denotation incorporated while leaving out reification, for now)
[6] bit.ly/1kXcPI2 -- Refication example
[7] bit.ly/1kXdwB5 -- Statements I made the other day about the World Cup.


Kingsley

Fixed Links:

[1] http://bit.ly/1cjYwqN -- DNS based Linked Computer Network (Internet)
[2] http://bit.ly/1e0pwdI -- HTTP based Linked Document Network (World Wide Web)
[3] http://bit.ly/INv6ag -- Linked Open Data Network (LOD Cloud)
[4] http://bit.ly/1lHD31h -- initial simplified Semantic Linked Open Data Network (Semantic Web, modulo denotation )
[5] http://bit.ly/1qRRYcn -- revised Semantic Linked Open Data Network (Semantic Web, with denotation incorporated while leaving out reification, for now)
[6] http://bit.ly/1kXcPI2 -- Refication example
[7] http://bit.ly/1kXdwB5 -- Statements I made the other day about the World Cup.

-- 

Regards,

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

smime.p7s
Description: S/MIME Cryptographic Signature

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)