Re: Use of owl:equivalentProperty with owl:AnnotationProperty

On 17 Jan 2017, at 13:43, Ignazio Palmisano <ipalmisano.mailings@gmail.com<mailto:ipalmisano.mailings@gmail.com>> wrote:

On 17 Jan 2017 13:38, "Svensson, Lars" <L.Svensson@dnb.de<mailto:L.Svensson@dnb.de>> wrote:
Dear all,

This is a newbie question, so please bear with me if the answer is obvious or has been answered elsewhere...

>From what I can see, the OWL Syntax does not allow axioms stating that two annotation properties are equal [1]. Is there a particular reason for that design decision?

Annotations are under-designed. We ran out of steam. What we have in there *at all* is a bit of a rump move.

My use case is to align two vocabularies and have found property pairs that are equivalent. I realise that I can use owl:subPropertyOf instead but the properties really are equivalent.


You can use two sub property axioms to achieve the same result,

Yep.

but the main issue is that annotation properties are ignored, to an extent, by reasoners.
I believe also that that might be the reason there is no equivalent axiom for annotation properties - they don't have the same semantics as the other properties.
I.


The key is what do you expect from this and from what tools?

If you are expecting an OWL reasoner to do anything useful with such an alignment…I’d be skeptical. They are free to ignore annotations when reasoning and generally do. I.e., just throw them away.

If *you* want to do something with annotations, I would focus on clearly defining what you want to have happen. E.g., if two annotation properties are aligned, should the axiom closure act as if all annotations are rewritten to use both properties? If so, yay! Spec it and write a tool to do it. I wouldn’t try to force OWL to do it “for free” (though Ignazio’s suggestion would be my favoured path).

Cheers,
Bijan.