Re: OWL DB, OWL UML, etc.

>Lately as I've been using and teaching OWL much more, I think I have 
>come to a better understanding of OWL vs. OWL DL and the real power 
>of having a reasoner-specific profile of OWL (i.e. OWL DL).  I've 
>also begun scratching around on some papers and talking to some 
>colleagues, and it becomes clear that there are other subsets of OWL 
>Full that might also be tremendously useful for other kinds of 
>tools. For example, it became clear that a fair subset of the OWL 
>expressiveness can be accounted for in the  calculus used by 
>relational database systems -- interestingly, although this is a 
>much less expressive language than OWL, some things which are 
>outside of OWL DL (particularly inverseFunctionalProperties on 
>datatype properties) are easily covered with that calculus. 
>Similarly, Guus had shown at one point (and it needs to be 
>revisited) that UML has a different set of restrictions than would 
>naturally be OWL DL.
>  So, this is mostly musing, but the question it brings up is whether 
>there is a common core of OWL that would like be in all of these 
>subsets,

Well, one place to start is to ask whether RDFS represents a common 
part of these subsets or not. I would guess that rdfs:subPropertyOf 
might be the only thing that one could make out a case against. Of 
course, the actual common subset might be larger in other ways.

Just a first thought.

>or whether it makes sense to think of OWL as what is now referred to 
>as OWL Full, and to think of all these subsets as specialized 
>profiles for particular kinds of applications (and the commonality 
>would provide at least some kinds of interoperability - esp. with 
>respect to editing, visualization, etc.)
>  Anyone working on any of the OWL-xxx subsets they'd care to 
>discuss? I admit to starting work on OWL DB,  which strikes me as an 
>important one if database integration is really going to be a major 
>application of OWL, and wonder if this is something that others are 
>playing with.
>  In short, I think it was Bijan who once commented that once OWL was 
>out there might be a cottage industry in creating special subsets 
>for particular application classes - seems like that might not be a 
>bad idea...

I agree.

Pat

>  -JH
>
>--
>Professor James Hendler 
>http://www.cs.umd.edu/users/hendler
>Director, Semantic Web and Agent Technologies	  301-405-2696
>Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
>Univ of Maryland, College Park, MD 20742	  240-277-3388 (Cell)


-- 
---------------------------------------------------------------------
IHMC	(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32501			(850)291 0667    cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes

Received on Tuesday, 11 May 2004 17:22:19 UTC