> > Who else would be able to make assertions about "your" notion of Brussels > vis-a-vis "some other notion of Brussels" with any more authority than your > own? >
The person doing the integration of your dataset with some other dataset for some purpose of *theirs*. The kind of correspondence needed is a function of the purpose of combining the data. Ignoring owl:sameAs statements isn't an option? > Not if you have a mixture of owl:sameAs statements you *have* to use and ones you *can't*. Whereas if you have x:correspondsTo, y:correspondsTo and z:correspondsTo, it *is* easy to say that y:correspondsTo (but not the other two) is the same as owl:sameAs.
