Re: [web-annotation] Clarify ability to reason with annotations - note additional statements to add to reasoner

-1 for several reasons.

Technically it is incorrect. A Motivation is an instance of a class, 
it is not a predicate. Thus inferring that a motivation is a predicate
 would be simply wrong. If we wanted to go that route, we would simply
 put that exact triple in the annotation: body motivation target
Further, we tried to do this in the past by having a predicate 
associated with the annotation, and the feedback was that it was both 
strange to have a non metadata reference to a predicate and did not 
meet the needs of the people that wanted it, as with the cardinality 
of body and target, it might be impossible (no body), or wrong 
(multiple heterogeneous bodies or targets)

Further, the actual inference logic even if it was reasonable, would 
be significantly more complex than that. It would need to take into 
account the intended semantics of the motivation (a comment body on an
 annotation with motivation highlighting does not highlight the 
target) and the existence of hasRole on the various resources.

>From a principles standpoint, we have tried hard to explicitly not 
talk about RDF related features and functionality in the model 
document to the point taking out "ontology" and other similar words. 
To start in the introduction to talk about reasoners and inferencing 
is completely counter to that stated principle.  We have #24 which 
even goes so far as to state the opposite of this proposal.

And editorially, introducing a somewhat normative statement in a note 
in the introduction before any of the concepts are described is just 
not the right solution, even if we were to accept that this is a real 
issue, against all previous discussion.

-- 
GitHub Notif of comment by azaroth42
See 
https://github.com/w3c/web-annotation/issues/98#issuecomment-153770875

Received on Wednesday, 4 November 2015 15:51:47 UTC