Re: Q: how to encode annotation "roles"

Jacco-

My understanding of your #4 approach---which I would argue for among
your suggestions--seems to run afoul of the  restriction of at most
one Body in the current proposal. I argued against that restriction in
the pre-public discussions as being needlessly closing the world.  The
counter argument was that allowing multiple Bodies can lead to
ambiguity in case there are also multiple Targets.  That is true, but
it is a generic problem about associations of one part of a graph with
another.  My position is that such ambiguity has several ways to
resolve, and which choice can best be left to communities of practice
or within an application.  Among these are the restriction now
mandatory, a formal(?) convention that within an Annotation all Bodies
apply to all Targets, use of rule languages that can be enforced
and/or validated by applications at Annotation generation time or
consumption time, use of domain-specific relations that might relate
Targets to Bodies, or agreement on formal semantics of a use of
dcterms:isReferencedBy. Probably there are others.

A use case I offered is an Annotation that needs to have Bodies
expressed in several natural languages (or for that matter, in several
different domain terminologies), even for a single Target. This seems
to me like a special case of your role use case, but in any case, the
other three of your solutions seem to require problem-specific
extensions to OA and OAX for no other reason than to solve a
reasonable problem of multiplicity.  I find them less favorable as a
part of the standard for much the same reason I don't like the unitary
oa:hasBody. As a recovering algebraist, I prefer generality over
restriction where-ever restriction gains you nothing but simplicity.
All the more so, when that simplicity can be easily gained by
communities of practice.

I'm hoping that the restriction to at most one Body will be
reconsidered.  Maybe it deserves its own thread.

Bob Morris

On Fri, May 25, 2012 at 8:13 AM, Jacco van Ossenbruggen
<Jacco.van.Ossenbruggen@cwi.nl> wrote:
> Dear all,
>
> I'm looking at the cookbook examples to learn how I could/should encode my
> current annotations in OA.
>
> I think the Eiffeltower example at [1] is pretty close to what I need.  But
> let's suppose I need a little more and want to explicitly assign different
> "roles" to my annotations, say the "depicted object" role (the Eiffel
> Tower), the "associated event" role (e.g. the 1889 World Fair) and the
> "associated person" role (e.g. Gustave Eiffel). Note that all three example
> tags have dbpedia entries and could be regarded as semantic tags in the oax
> sense.
>
> I can see many different ways to make such roles explicit in the OA model:
>
> 1: Creating subclasses of oa:Annotation or oax:Description for each role
> 2: Creating subproperties of oax:hasSemanticTag for each role
> 3: Creating a new property on the oa:Annotation instance defining the role
> 4: Add a hasBody with a property on the body defining the role
>
> There are undoubtedly other patterns I've overlooked.   Is there any best
> practice in this community about how to encode this?
> I'm not sure whether "roles" is the best term, but I'm quite convinced this
> is a common pattern. What I need is a mapping from simple triples like
>
> <http://alturl.com/xxbxnn> ns1:depictedObject
> <http://dbpedia.org/page/Eiffel_Tower>;
> <http://alturl.com/xxbxnn> ns1:associatedEvent
> <http://dbpedia.org/page/Exposition_Universelle_(1889)>;
> <http://alturl.com/xxbxnn> ns1:associatedPerson
> <http://dbpedia.org/page/Eiffel_Gustave_Eiffel>;
>
> to the OA model.
>
> Any help would be appreciated,
>
> Regards, Jacco
>
> [1]
> http://www.w3.org/community/openannotation/wiki/SE_Semantically_Tagging_an_Image
>



-- 
Robert A. Morris

Emeritus Professor  of Computer Science
UMASS-Boston
100 Morrissey Blvd
Boston, MA 02125-3390

IT Staff
Filtered Push Project
Harvard University Herbaria
Harvard University

email: morris.bob@gmail.com
web: http://efg.cs.umb.edu/
web: http://etaxonomy.org/mw/FilteredPush
http://www.cs.umb.edu/~ram
===
The content of this communication is made entirely on my
own behalf and in no way should be deemed to express
official positions of The University of Massachusetts at Boston or
Harvard University.

Received on Thursday, 31 May 2012 02:52:20 UTC