Re: roles and multiple bodies

I'm fine with a "Proxy" like construction (x stands for y, in the context
of z, with these additional properties).  We avoided that for specific
resources as it was seen to be overly complex in favor of the direct tree
structure, but it does mean that implementers can ignore the
property/object more easily.

R


On Wed, Aug 5, 2015 at 7:19 AM, Paolo Ciccarese <paolo.ciccarese@gmail.com>
wrote:

> Thank you Ivan for better articulating my short comment :)
>
> Blank nodes make some good sense there as the assignment is very localized
> by nature.
> Plus I really like the possibility of ignoring such complexity.
>
> Although, It is clear that in JSON terms this solution requires some
> 'stitching'...
>
> On Wed, Aug 5, 2015 at 3:19 AM, Ivan Herman <ivan@w3.org> wrote:
>
>>
>> > On 04 Aug 2015, at 23:16 , Denenberg, Ray <rden@loc.gov> wrote:
>> >
>> > Hi Paolo,
>> >
>> > I felt the forAnnotation to be necessary, because without it, you would
>> be saying “this role is assigned to this body”, independent of any
>> particular annotation. So if some other role were to be assigned to that
>> body, for a different annotation, there would be a conflict.
>>
>> I do not think there is a problem. If you have
>>
>> <anno>
>>         oa:hasBody <uri1>;
>>         oa:hasBody <uri2>;
>>         oa:hasRoleAssignment [
>>                 oa:role: "role1",
>>                 oa:assignedTo: <uri1>
>>         ]
>>         …
>>
>> what we say, in human terms, that the subject of a triple with the
>> predicate hasRoleAssignment is an annotation, and its object is an
>> 'Assignment', and the predicates refer back to the subject of the
>> hasRoleAssignment. Yes, it is an application specific semantics, but that
>> we have in any case. (Actually, if we really want to get into formal
>> semantics, the OWL2 Property Chain may express that kind of chaining if we
>> really care… most users won't.) I would not worry whether the object of
>> hasRoleAssignment is a blank node or not, b.t.w. For me, using a blank node
>> at that point is perfectly fine.
>>
>> Bottom line: yes that is a solution. Whether it is better or worse than
>> Rob's approach: I am not sure. I have a mild preference for this one,
>> because it cleanly separates the bodies from the roles, and if somebody
>> does not care about the roles than all this complication can be ignored.
>>
>>
>> Ivan
>>
>> P.S. (Why are blank nodes fine here? Because I do not think there is any
>> reasonable usage of merging to that particular object's triples with a
>> similar set of triples coming from a very different annotation, for
>> example. Blank nodes become an obstacle if the intention is to cleanly
>> merge, or refer to, a set of triples but, as long as they are, so to say,
>> localized, it is o.k. imho.)
>>
>>
>>
>> >
>> > Ray
>> >
>> > From: Paolo Ciccarese [mailto:paolo.ciccarese@gmail.com]
>> > Sent: Tuesday, August 04, 2015 5:13 PM
>> > To: Denenberg, Ray
>> > Cc: Web Annotation
>> > Subject: Re: roles and multiple bodies
>> >
>> > Hi Ray,
>> > Is the subject of oa:hasRoleAssignment still the annotation?
>> > I would probably  remove oa:forAnnotation as it looks redundant to me.
>> >
>> > It is a little bit indirect in JSON but it seems a possible solution.
>> > Who needs to be more specific could add those sections.
>> >
>> > Paolo
>> >
>> > On Tue, Aug 4, 2015 at 4:55 PM, Denenberg, Ray <rden@loc.gov> wrote:
>> > Has the following approach been considered?  I don’t think it violates
>> any of the principles that Rob is defending, because it explicitly narrows
>> the scope of the role to the annotation:
>> >
>> >
>> > <http://www.example.org/annotation>      a  oa:Annotation ;
>> >
>> >           oa:hasTarget  <http://www.example.org/target>      ;
>> >
>> >           oa:hasBody  <http://www.example.org/body1>  ;
>> >           oa:hasBody  <http://www.example.org/body2>   ;
>> >           oa:hasBody  <http://www.example.org/body2>   ;
>> >
>> > oa:hasRoleAssignment [
>> >                       oa:role                   <role1>   ;
>> >                        oa:assignedTo     <http://www.example.org/body1>
>> ;
>> >                       oa:forAnnotation  <
>> http://www.example.org/annotation>   ] ;
>> >
>> > oa:hasRoleAssignment [
>> >                       oa:role                   <role2>   ;
>> >                        oa:assignedTo     <http://www.example.org/body2>
>> ;
>> >                       oa:forAnnotation  <
>> http://www.example.org/annotation>   ] ;
>> >
>> >
>> > oa:hasRoleAssignment [
>> >                       oa:role                   <role3>   ;
>> >                        oa:assignedTo     <http://www.example.org/body3>
>> ;
>> >                       oa:forAnnotation  <
>> http://www.example.org/annotation>   ] .
>> > --Ray
>> >
>> >
>> >
>> > --
>> > Dr. Paolo Ciccarese
>> > Principal Knowledge and Software Engineer at PerkinElmer Innovation Lab
>> > Assistant Professor of Neurology at Harvard Medical School
>> > Assistant in Neuroscience at Mass General Hospital
>> > ORCID: http://orcid.org/0000-0002-5156-2703
>>
>>
>> ----
>> Ivan Herman, W3C
>> Digital Publishing Activity Lead
>> Home: http://www.w3.org/People/Ivan/
>> mobile: +31-641044153
>> ORCID ID: http://orcid.org/0000-0003-0782-2704
>>
>>
>>
>>
>>
>
>
> --
> Dr. Paolo Ciccarese
> Principal Knowledge and Software Engineer at PerkinElmer Innovation Lab
> Assistant Professor of Neurology at Harvard Medical School
>
> Assistant in Neuroscience at Mass General Hospital
> ORCID: http://orcid.org/0000-0002-5156-2703
>



-- 
Rob Sanderson
Information Standards Advocate
Digital Library Systems and Services
Stanford, CA 94305