- From: Bijan Parsia <bparsia@cs.man.ac.uk>
- Date: Fri, 30 May 2008 02:22:52 +0100
- To: OWL Working Group WG <public-owl-wg@w3.org>
On May 28, 2008, at 7:32 PM, Alan Ruttenberg wrote: [snip] >> I would also like some acknowledgement that some systems do >> "auto-imports", where they perform imports based on the URIs used >> in the >> document (especially as names for classes and properties), even >> when no >> imports directive is present. The only system I know of that does that is CWM, and then only as an option. Are there others? >> We don't need to pass judgement on this >> practice, but if we don't mention it, it's rather confusing how this >> (perfectly legitimate) practice relates to OWL2 import. > > Legitimate in what sense? I wouldn't expect an OWL processor to do > this. > I'm with Peter on this one. Say nothing, because what can be said? Or we could recommend against it, at least as part of "normal" OWL processing. Indeed, it's hard to see how it's legitimate as a form of imports or as having any relation to the actual meaning of the document. (I know Tim has this thing about it, but he's never been very convincing.) I don't see it as any more discussion worthy than saying that a system can collect data by spidering in general, or by seeAlso following, or by using it's own cache of background knowledge, or by automapping into a foundational ontology. These are all legitimate *knowledge acquisition* practices, but have nothing to do with the specced meaning of an OWL document. Hmm. Now that I wrote that, I guess I would be fine with there being some text (perhaps in the primer?) distinguishing imports from KA. But is there really any *general* confusion about this? I mean, I know *Tim* is confused about this, and is happy to confuse others, but is there any evidence that the general set of users are so confused? I see no evidence to this effect. (Back in early OWL days, there was a lot of confusion, rolled over from DAML+OIL, about namespaces and imports, but that seems long gone.) Cheers, Bijan.
Received on Friday, 30 May 2008 01:23:34 UTC