Use of sioc:sibling for users

[Alexandre Passant]

Hi all,

sioc:sibling is currently defined with sioc:Item as both domain and range.
I think that's a useful property to model sibling user accounts, e.g.
creating a direct relation between a Twitter and FB account, w/o
having to rely on a foaf:Person to link both.

Any opinions? Could we remove the domain / range restrictions of
sioc:sibling to allow such use-cases?

Best,

Alex.

--
Dr. Alexandre Passant - @terraces
Founder, CEO - http://seevl.net - @seevl
http://facebook.com/seevl.net - http://angel.co/seevl

[Olivier GENDRIN]

On Fri, Mar 2, 2012 at 10:52 AM, Alexandre Passant <al...@seevl.net> wrote:
> sioc:sibling is currently defined with sioc:Item as both domain and range.
> I think that's a useful property to model sibling user accounts, e.g.
> creating a direct relation between a Twitter and FB account, w/o
> having to rely on a foaf:Person to link both.
>
> Any opinions? Could we remove the domain / range restrictions of
> sioc:sibling to allow such use-cases?

You should have at least a foaf:agent handling both accounts. It can
be anonymous.

--
Olivier G.
http://identi.ca/lespacedunmatin
http://www.lespacedunmatin.info/blog/

[Alexandre Passant]

On Mar 2, 10:03 am, Olivier GENDRIN <olivier.gend...@gmail.com> wrote:

> On Fri, Mar 2, 2012 at 10:52 AM, Alexandre Passant <a...@seevl.net> wrote:
> > sioc:sibling is currently defined with sioc:Item as both domain and range.
> > I think that's a useful property to model sibling user accounts, e.g.
> > creating a direct relation between a Twitter and FB account, w/o
> > having to rely on a foaf:Person to link both.
>
> > Any opinions? Could we remove the domain / range restrictions of
> > sioc:sibling to allow such use-cases?
>
> You should have at least a foaf:agent handling both accounts. It can
> be anonymous.

That's my concern. Why should I create an additional triple instead of
a direct relation ?

<http://twitter.com/foo> sioc:sibling <http://facebook.com/foo>

vs

[] foaf:account <http://twitter.com/foo> ; <http://facebook.com/foo> .

Which also leads to

SELECT sibling WHERE { <http:twitter.com/foo> sioc:sibling ?sibing }

vs

SELECT sibling WHERE { [] sioc:sibling <http:twitter.com/foo> ;
sioc:sibling ?sibing . FILTER (?sibling != <http:twitter.com/foo> }

Best,

Alex.

>
> --
> Olivier G.http://identi.ca/lespacedunmatinhttp://www.lespacedunmatin.info/blog/

[Olivier GENDRIN]

On Fri, Mar 2, 2012 at 10:52 AM, Alexandre Passant <al...@seevl.net> wrote:
> sioc:sibling is currently defined with sioc:Item as both domain and range.
> I think that's a useful property to model sibling user accounts, e.g.
> creating a direct relation between a Twitter and FB account, w/o
> having to rely on a foaf:Person to link both.
>
> Any opinions? Could we remove the domain / range restrictions of
> sioc:sibling to allow such use-cases?

Having read the definition, I’m not shure you can use sibling to link
a Twitter and a FaceBook account : that would suppose that the
contents of both are the same, wich is rarely the case.

[Alexandre Passant]

On Fri, Mar 2, 2012 at 10:26 AM, Olivier GENDRIN
<olivier...@gmail.com> wrote:
> On Fri, Mar 2, 2012 at 10:52 AM, Alexandre Passant <al...@seevl.net> wrote:
>> sioc:sibling is currently defined with sioc:Item as both domain and range.
>> I think that's a useful property to model sibling user accounts, e.g.
>> creating a direct relation between a Twitter and FB account, w/o
>> having to rely on a foaf:Person to link both.
>>
>> Any opinions? Could we remove the domain / range restrictions of
>> sioc:sibling to allow such use-cases?
>
> Having read the definition, I’m not shure you can use sibling to link
> a Twitter and a FaceBook account : that would suppose that the
> contents of both are the same, wich is rarely the case.

Indeed - good point.
So maybe edit the definition (I find the naming of sibling appropriate
for this), or create a new property?

Alex.

>
> --
> Olivier G.
> http://identi.ca/lespacedunmatin
> http://www.lespacedunmatin.info/blog/
>

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