Comments About SIOC concepts / properties

On 7/19/06, John Breslin <john.breslin@deri.org> wrote:

>  > 2) sioc:Comment
>
> This would need to be moved to a types module, to keep the core SIOC

> ontology simple.  We can then add other subtypes of Post later like
> Sticky, Announcement, etc.

Attaching here a nice drawing by [GNU] showing proposed relations between SIOC, SKOS and IBIS.
The IBIS part formalizes the part about SIOC and argumented discussions mentioned earlier.

Not sure if [GNU] has a particular vocabulary in mind, but there is one made by Danny Ayers [1].

[1] http://www.schemaweb.info/schema/SchemaDetails.aspx?id=4



Regarding how to express Comments in SIOC:

The best place to express different types of posts (comments, ...) as well as different types of other SIOC concepts (forums, etc.) is in modules outside the core SIOC ontology. That helps the ontology stay clean yet flexible and extensible.


With this clarification, now there is a question - what do we need SKOS for in this situation? There is already a tool in RDFS to define hierarchies - rdfs:subClassOf - is there something it does not allow to do that SKOS does? SKOS would be useful to express category hierarchies for online community sites, but I do not see the value added by it in this particular case.


In terms of implementation we could define the Post types module as a hierarchy of subclasses of SIOC:Post. To held existing consumers of RDF data that maybe are not capable of reasoning over schemas we can define both classes explicitly in the export of SIOC data for now:

[ <post> a sioc:Post .
  <post> a sioc_type:Comment . ]

If it is true that [ sioc_type:Comment rdfs:subClassOf sioc:Type . ] then in the future we will be able to drop the line [ <post> a sioc:Post . ] and consumers will infer this from knowing it's a sioc_type:Comment (although then it looks a bit awkward to have to use a different namespace when defining a Comment).


Note: RDF allows to assign a number of types at the same time - there are no problems to define a "Sticky + Announcement" or "Sticky + Comment" using the terms John mentions above.

Uldis


[ http://captsolo.net/info/ ]



--~--~---------~--~----~------------~-------~--~----~

You received this message because you are subscribed to the Google Groups "SIOC-Dev" group.
To post to this group, send email to sioc-dev@googlegroups.com
To unsubscribe from this group, send email to sioc-dev-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sioc-dev
-~----------~----~----~----~------~----~------~--~---


Comments About SIOC concepts / properties

Great drawing!

The IBIS idea is an important one, worthy of addition to SIOC. The
latest information on IBIS is based on the dialog mapping platform
created by Jeff Conklin, now found in the Compendium tool [1] , and at
Jeff's website [2]. It is also worth noting that the Knowledge Media
Institute (KMi) [3] is leading research on frameworks that include
Compendium. I have explored combinations of technologies not unlike
the combinations of SIOC; I call that "augmented storytelling" [4]. In
place of SKOS, which didn't exist when I started looking (2002), I
used topic maps, which were the subject of an Addison-Wesley book I
had edited and just published. Today, we are evolving "subject maps"
as a more powerful system of representation of subjects, some of which
require complex subject identity representations.

Subject identity turns out to be valuable in cases where ontologies
are to be merged in ways which do not include the necessity of
supporting "authoritative question answering." Patrick Durusau and I
gave a teleconference to the Ontolog community [5], in which we
described a process we call "subject-centric merging", an idea that
might augment the discussion at the wiki [6]. Slides and an mp3 of the
talk are available at [5]. I would be happy to participate in
discussions about these ideas.

Jack
[1] http://compendiuminstitute.org/
[2] http://cognexus.org/
[3] http://kmi.open.ac.uk/
[4] http://www.nexist.org/nsc2004/
[5] http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2006_04_27
[6] http://esw.w3.org/topic/SIOC/CombiningOntologies

On 7/19/06, Uldis Bojars wrote:
> On 7/19/06, John Breslin wrote:
>
> > > 2) sioc:Comment
> >
> > This would need to be moved to a types module, to keep the core SIOC
> > ontology simple. We can then add other subtypes of Post later like
> > Sticky, Announcement, etc.
>
> Attaching here a nice drawing by [GNU] showing proposed relations between
> SIOC, SKOS and IBIS.
> The IBIS part formalizes the part about SIOC and argumented discussions
> mentioned earlier.
> Not sure if [GNU] has a particular vocabulary in mind, but there is one made
> by Danny Ayers [1].
>
> [1]
> http://www.schemaweb.info/schema/SchemaDetails.aspx?id=4
>
>
> Regarding how to express Comments in SIOC:
>
> The best place to express different types of posts (comments, ...) as well
> as different types of other SIOC concepts (forums, etc.) is in modules
> outside the core SIOC ontology. That helps the ontology stay clean yet
> flexible and extensible.
>
> With this clarification, now there is a question - what do we need SKOS for
> in this situation? There is already a tool in RDFS to define hierarchies -
> rdfs:subClassOf - is there something it does not allow to do that SKOS does?
> SKOS would be useful to express category hierarchies for online community
> sites, but I do not see the value added by it in this particular case.
>
> In terms of implementation we could define the Post types module as a
> hierarchy of subclasses of SIOC:Post. To held existing consumers of RDF data
> that maybe are not capable of reasoning over schemas we can define both
> classes explicitly in the export of SIOC data for now:
> [
a sioc:Post .
>
a sioc_type:Comment . ]
>
> If it is true that [ sioc_type:Comment rdfs:subClassOf sioc:Type . ] then in
> the future we will be able to drop the line [
a sioc:Post . ] and
> consumers will infer this from knowing it's a sioc_type:Comment (although
> then it looks a bit awkward to have to use a different namespace when
> defining a Comment).
>
> Note: RDF allows to assign a number of types at the same time - there are no
> problems to define a "Sticky + Announcement" or "Sticky + Comment" using the
> terms John mentions above.
>
> Uldis
>
> [ http://captsolo.net/info/ ]
>
> >
>
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "SIOC-Dev" group.
To post to this group, send email to sioc-dev@googlegroups.com
To unsubscribe from this group, send email to sioc-dev-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sioc-dev
-~----------~----~----~----~------~----~------~--~---