Hubjects and Open GUID diagram
11 views
Skip to first unread message
bernard
Sep 26, 2008, 3:20:01 AM9/26/08
to Open GUID Discussion
I just uploaded a very simple diagram under Files/openguid.pdf (can't
figure how to attach it to the message) showing my current
understanding of how Open GUID is an implementation of my old hubject
notion, going live through "anchoring" in both Web representations and
human language.
Comments welcome. Could be improved and used for OpenGUID presentation/
outreach ...
figure how to attach it to the message) showing my current
understanding of how Open GUID is an implementation of my old hubject
notion, going live through "anchoring" in both Web representations and
human language.
Comments welcome. Could be improved and used for OpenGUID presentation/
outreach ...
ja...@openguid.net
Sep 26, 2008, 4:43:34 PM9/26/08
to Open GUID Discussion
Sweet, thanks Bernard.
Diagrams would be great for the documentation part of the site.
Although in this case I would need to explain or reference the term
'hubject'. Do you have any copy that explain the concept succinctly?
Or a quick ref on your site?
Diagrams would be great for the documentation part of the site.
Although in this case I would need to explain or reference the term
'hubject'. Do you have any copy that explain the concept succinctly?
Or a quick ref on your site?
Bernard Vatant
Sep 29, 2008, 5:02:27 AM9/29/08
to open-guid-...@googlegroups.com
Hi Jason
Glad you like itSweet, thanks Bernard.
Diagrams would be great for the documentation part of the site. Although in this case I would need to explain or reference the term 'hubject'. Do you have any copy that explain the concept succinctly? Or a quick ref on your site?
Actually, I'm not sure this "hubject" term needs to survive at all.
It's a term I coined three years ago and throw on line to feed
reflexion on the issue.
If you, and the open GUID community find it relevant, it's open vocabulary. But as said before, my thoughts about it have evolved since, and not sure that pointing to existing resources would help clarifying. If I had to give a definition today, it will be a semiotic one as the one below.
Hubject (neologism, 2005, from hub and subject)
A technical implementation of a link between two or more signs (or representations), agreed to have the same referent (representing the same subject). If the signs linked through a hubject are defined in a formal language, the semantics of the linked signs may or may not be consistent with each other, they might convey identical or different significations (views) of the common referent. A hubject has no other declared semantics than the link it provides.
In the Semantic Web, signs linked by hubjects are resources signified by URIs. Having equivalent semantics declared in a Semantic Web language, using constructions such as owl:sameAs or owl:equivalentClass, can be seen as a sufficient condition to be linked to the same hubject, but it is not necessary.
Actually, since Open GUID defines URIs, one can ask the nature and type of the resource identified by such URIs. Hubject is a good candidate for that.
So I would have no problem with introducing the class oguid:Hubject, and have the following declarations
http://openguid.net/e6bfecbd-da25-102b-9a03-2db401e887ec rdf:type oguid:Hubject
BTW, one thing not very clear in the current spec, is how the above URI is itself linked to the identical URIs it relates
I find today, dereferencing the URI
http://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://wordnet.princeton.edu/synset/109444100
And soon
http://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://dbpedia.org/page/Star
So there is some recursivity in a oguid URI definition. It links identical URIs, including itself. Right?
Bernard
If you, and the open GUID community find it relevant, it's open vocabulary. But as said before, my thoughts about it have evolved since, and not sure that pointing to existing resources would help clarifying. If I had to give a definition today, it will be a semiotic one as the one below.
Hubject (neologism, 2005, from hub and subject)
A technical implementation of a link between two or more signs (or representations), agreed to have the same referent (representing the same subject). If the signs linked through a hubject are defined in a formal language, the semantics of the linked signs may or may not be consistent with each other, they might convey identical or different significations (views) of the common referent. A hubject has no other declared semantics than the link it provides.
In the Semantic Web, signs linked by hubjects are resources signified by URIs. Having equivalent semantics declared in a Semantic Web language, using constructions such as owl:sameAs or owl:equivalentClass, can be seen as a sufficient condition to be linked to the same hubject, but it is not necessary.
Actually, since Open GUID defines URIs, one can ask the nature and type of the resource identified by such URIs. Hubject is a good candidate for that.
So I would have no problem with introducing the class oguid:Hubject, and have the following declarations
http://openguid.net/e6bfecbd-da25-102b-9a03-2db401e887ec rdf:type oguid:Hubject
BTW, one thing not very clear in the current spec, is how the above URI is itself linked to the identical URIs it relates
I find today, dereferencing the URI
http://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://wordnet.princeton.edu/synset/109444100
And soon
http://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://dbpedia.org/page/Star
So there is some recursivity in a oguid URI definition. It links identical URIs, including itself. Right?
Bernard
On Sep 26, 1:20 am, bernard <bernard.vat...@mondeca.com> wrote:I just uploaded a very simple diagram under Files/openguid.pdf (can't figure how to attach it to the message) showing my current understanding of how Open GUID is an implementation of my old hubject notion, going live through "anchoring" in both Web representations and human language. Comments welcome. Could be improved and used for OpenGUID presentation/ outreach ...
--
Bernard Vatant
Senior
Consultant
Vocabulary & Data Engineering
Tel: +33 (0) 971 488 459
Mail: bernard...@mondeca.com
----------------------------------------------------
Mondeca
3, cité Nollez 75018 Paris France
Web: www.mondeca.com
Blog: Leçons
de Choses
----------------------------------------------------
ja...@openguid.net
Sep 29, 2008, 9:50:38 PM9/29/08
to Open GUID Discussion
Hmm, that is interesting to have a class defined for Open GUIDs.
Although without properties I'm not sure if anyone will get value from
it.
Yes, there will be an RDF file available listing all the known
identical statements. It will contain links from the Open GUID to
each resource declared identical. It will also relate Open GUIDs that
have been determined duplicate. I think with the transitive and
symmetric behavior spelled out in the spec, I would avoid excess
identical statements to itself or between each related URI in the
repository.
Though that does bring up the question of whether there is value of
maintaining a repository of all things marked identical, even if there
is no Open GUID defined. I'm not sure.
On Sep 29, 3:02 am, Bernard Vatant <bernard.vat...@mondeca.com> wrote:
> Hi JasonSweet, thanks Bernard.Glad you like itDiagrams would be great for the documentation part of the site. Although in this case I would need to explain or reference the term 'hubject'. Do you have any copy that explain the concept succinctly? Or a quick ref on your site?Actually, I'm not sure this "hubject" term needs to survive at all. It's a term I coined three years ago and throw on line to feed reflexion on the issue.
> And soonhttp://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://dbpedia.org/page/Star
>
> BernardVatantSeniorConsultantVocabulary& Data Engineering
> ----------------------------------------------------Mondeca3, citéNollez75018 Paris France
Although without properties I'm not sure if anyone will get value from
it.
Yes, there will be an RDF file available listing all the known
identical statements. It will contain links from the Open GUID to
each resource declared identical. It will also relate Open GUIDs that
have been determined duplicate. I think with the transitive and
symmetric behavior spelled out in the spec, I would avoid excess
identical statements to itself or between each related URI in the
repository.
Though that does bring up the question of whether there is value of
maintaining a repository of all things marked identical, even if there
is no Open GUID defined. I'm not sure.
On Sep 29, 3:02 am, Bernard Vatant <bernard.vat...@mondeca.com> wrote:
> Hi JasonSweet, thanks Bernard.Glad you like itDiagrams would be great for the documentation part of the site. Although in this case I would need to explain or reference the term 'hubject'. Do you have any copy that explain the concept succinctly? Or a quick ref on your site?Actually, I'm not sure this "hubject" term needs to survive at all. It's a term I coined three years ago and throw on line to feed reflexion on the issue.
> If you, and the open GUID community find it relevant, it's open vocabulary. But as said before, my thoughts about it have evolved since, and not sure that pointing to existing resources would help clarifying. If I had to give a definition today, it will be a semiotic one as the one below.
> Hubject(neologism, 2005, fromhubandsubject)A technical implementation of a link between two or more signs (or representations), agreed to have the same referent (representing the same subject). If the signs linked through a hubject are defined in a formal language, the semantics of the linked signs may or may not be consistent with each other, they might convey identical or different significations (views) of the common referent. A hubject has no other declared semantics than the link it provides.
> In the Semantic Web, signs linked by hubjects are resources signified by URIs. Having equivalent semantics declared in a Semantic Web language, using constructions such as owl:sameAs or owl:equivalentClass, can be seen as a sufficient condition to be linked to the same hubject, but it is not necessary.
> Actually, since Open GUID defines URIs, one can ask the nature and type of the resource identified by such URIs. Hubject is a good candidate for that.
> So I would have no problem with introducing the class oguid:Hubject, and have the following declarationshttp://openguid.net/e6bfecbd-da25-102b-9a03-2db401e887ec rdf:type oguid:Hubject
> Actually, since Open GUID defines URIs, one can ask the nature and type of the resource identified by such URIs. Hubject is a good candidate for that.
> BTW, one thing not very clear in the current spec, is how the above URI is itself linked to the identical URIs it relates
> I find today, dereferencing the URIhttp://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://wordnet.princeton.edu/synset/109444100
> And soonhttp://openguid.net/e6bfe3c9-da25-102b-9a03-2db401e887ec oguid:identical http://dbpedia.org/page/Star
> So there is some recursivity in a oguid URI definition. It links identical URIs, including itself. Right?
> BernardOn Sep 26, 1:20 am, bernard<bernard.vat...@mondeca.com>wrote:I just uploaded a very simple diagram under Files/openguid.pdf (can't figure how to attach it to the message) showing my current understanding of how Open GUID is an implementation of my old hubject notion, going live through "anchoring" in both Web representations and human language. Comments welcome. Could be improved and used for OpenGUID presentation/ outreach ...--
>
> BernardVatantSeniorConsultantVocabulary& Data Engineering
> ----------------------------------------------------Mondeca3, citéNollez75018 Paris France
Reply all
Reply to author
Forward
0 new messages