- From: Bob Morris <morris.bob@gmail.com>
- Date: Fri, 4 Dec 2009 13:50:21 -0500
- To: Pat Hayes <phayes@ihmc.us>
- Cc: Peter DeVries <pete.devries@gmail.com>, public-lod@w3.org, dmozzherin@gmail.com
On Fri, Dec 4, 2009 at 1:02 PM, Pat Hayes <phayes@ihmc.us> wrote: > > On Dec 4, 2009, at 11:14 AM, Bob Morris wrote: > >>> In any civilized ontology language, a class is an instance. Use OWL 2 >>> from >>> now on. >>> Pat >> >> Ah, a nice idea, but I doubt it will accomplish what Pete is trying to >> do (Or my limited understanding of OWL 2 is wrong---which is a good >> hypothesis). >> http://www.w3.org/TR/2009/REC-owl2-syntax-20091027/#Metamodeling >> says that when a resource is mentioned as both an individual and a >> class, the two mentions are interpreted independently. > > Well, true, so its not perfect. But it is better than OWL 1 in this regard. Amen. In fact, my first impression of OWL 2 is that one might think of it as an answer to the question: How can we robustly relax the restriction in OWL1 that we made too cavalierly and that interfere with important use cases. As I recall, the W3 docs even say that somewhere. One concern I am still trying to wrap my head around in biodiversity applications---and maybe in lots of distributed science data applications---is whether both the RL and QL profiles are needed in the same ontology and in ways that their intersection is insufficiently expressive. That's not for the question at hand though (I think). As to the perennial class vs instances question, I'd be interested to learn what kinds of scalability experiments have been done on the OWL2 EL profile (e.g. how many triples per class description in realistic cases.) Per another posting, a comprehensive taxonomic ontology probably requires at least three times, and more likely 10-20 times, more classes than the 357K classes reported for SNOMED CT in http://dx.doi.org/10.1016/j.artmed.2006.12.003. The number of distinct names is probably in similar ratio, as there are about 11 million names in the uBio NameBank http://www.ubio.org/ and about 1M in SNOMED Robert A. Morris Professor of Computer Science (nominally retired) UMASS-Boston 100 Morrissey Blvd Boston, MA 02125-3390 Associate, Harvard University Herbaria email: ram@cs.umb.edu web: http://bdei.cs.umb.edu/ web: http://etaxonomy.org/FilteredPush http://www.cs.umb.edu/~ram phone (+1)617 287 6466
Received on Friday, 4 December 2009 18:50:57 UTC