W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > September 2007

RE: One ontology schema - heterogeneous instance bases

From: Michel_Dumontier <Michel_Dumontier@carleton.ca>
Date: Mon, 10 Sep 2007 10:34:01 -0400
To: wangxiao@musc.edu, Eric Jain <Eric.Jain@isb-sib.ch>
Cc: public-semweb-lifesci@w3.org
Message-id: <AB349814F1ECB143A5D4CD29C7A6456901B83CFD@CCSEXB10.CUNET.CARLETON.CA>


> Can't speak for BioPAX, but I can say that for me one problem with DC 
> is that it's RDF(S), so I can't use the properties in my OWL 
> restrictions etc!
>
> There are some OWL versions (Protege even allows you to import one), 
> but I don't know if that is a good idea, as far as I can see this 
> isn't official?
The protégé owl DC document [1] defines DC URIs as OWL Annotation Properties. Since annotation properties have no semantics, one won't be able to apply restrictions or reason about them. Redefining the DC terms as datatypes will lead to OWL Full upon integration of two different documents, which may not be handled by DL reasoners. Ideally, we would have an OWL ontology that defines the various terms (i.e. the class of Author), and makes object properties out of DC terms. However, this redefinition exemplifies the need for another URI. A challenge is which URIs to mint for such a document, and moreover, what URIs shall we use for the various authors. Unlike PubMed articles, I am not aware of a system to generate consistent and unique URIs for authors.


-=Michel=-

[1] http://protege.stanford.edu/plugins/owl/dc/protege-dc.owl
Received on Monday, 10 September 2007 14:35:02 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 10 December 2014 20:09:36 UTC