Issue 10: Provenance Mapping

For context, please see:
    https://github.com/w3c/web-annotation/issues/10

The issue brought up is that the provenance mapping [1]  may be internally
inconsistent.

The intent of the mapping was:

There is an document, annoDocument1, which was generated by the serializing
event. That event takes place at datetime2, and is performed by agent2.
The document is derivedFrom the conceptual annotation.
The annotation, anno1, was generated by an annotating event.  That event
takes place at datetim1, and is performed by agent1.

Thus oa:annotatedAt is the same as prov:generatedAt for the annotation
concept, and serializedAt is a shortcut to avoid minting another URI for
the annotation document.

Luc wonders if the generation-generation-ordering constraint [2] makes this
invalid.
I'm not sure that it applies here, as there are two entities involved --
the document and the conceptual annotation, each (in this case) with
different generation times, activities and agents. In the model we don't
mint a URI for the document, otherwise we could use prov natively and avoid
the oa:serializedAt and oa:serializedBy shortcuts.

Thoughts?

Rob


[1]: http://www.openannotation.org/spec/core/images/provmapping.png
[2]:
http://www.w3.org/TR/prov-constraints/#generation-generation-ordering_text

-- 
Rob Sanderson
Technology Collaboration Facilitator
Digital Library Systems and Services
Stanford, CA 94305