Re: symmetry of inverseOf

Michel Leonard Goldstein wrote:
> In order to decrease the amount of different property types in my ontology,
> I have made all properties that connect something to a paper to be called
> "ofPaper". They can be disambiguated and queried based on the class of the
> "source" (domain). This makes special sense for the system I'm building,
> because it always starts a search by a class type. On the other hand, the
> outgoing properties are named differently for each different class (for
> example, for the author of the paper there is a property called "hasAuthor",
> for the title, "hasTitle", and so on).
> In this case, the "ofPaper" property is the inverse of the "hasAuthor"
> property, but the symmetric assertion is not valid. If I understood the
> defined properties for OWL, there is no defined property that allows me to
> represent this unidirectional relationship. In the inverseOf property, there
> is a "vice versa" attached to the definition. Am I right? If so, why wasn't
> this considered interesting enough to be added to the definition?

My two cents:
If I understand your modelling, you could introduce a parent property 
for "hasAuthor", "hasTitle", etc. called for instance "hasAttribute" and 
this property could be declared to be the inverseOf "forPaper". As a 
result the implication would be a one way implication for the sub 
properties of "hasAttribute":

X hasAuthor Y => X hasAttribute Y => Y forPaper X

while you only have:

Y forPaper X => X hasAttribute

Best,

Fabien

-- 
"The process which led from the amoeba to man appeared
  to philosophers to obviously be progress,
  though whether the amoeba would agree with this opinion
  is not known."
                                     -- Bertrand Russell.
  ____________
|__ _ |_  http://www-sop.inria.fr/acacia/personnel/Fabien.Gandon/
|  (_||_) INRIA Sophia Antipolis - ph# (33)(0)4 92 38 77 88

Received on Thursday, 11 March 2004 11:42:20 UTC