RE: MT RDFS closure rule bug?

>  >
>>  >Pat,
>>  >
>>  >don't we need RDFS closure rules that add range and domain constraints
>>  >e.g.
>>  >
>>  >aaa [rdfs:range] yyy
>>  >yyy [rdfs:subClassOf] zzz
>>  >
>>  >then add
>>  >
>>  >aaa [rdfs:range] zzz
>>  >
>>  >
>>  >and similarly for rdfs:domain.
>>
>>  NO. That would be disastrous for the datatyping and in any case not
>>  make sense. Why do want them?
>>
>>
>
>We don't *want* them, they are just true!
>Or maybe I've been talking to Peter too much!
>
>Any interpretation of any
>
>>  >aaa [rdfs:range] yyy
>>  >yyy [rdfs:subClassOf] zzz
>
>
>is an interpretation of
>
>>  >aaa [rdfs:range] zzz
>
>
>thus the closure rule holds.
>
>(Not) Proof:
>
>Ahh, it depends on rdfs:range not being in the domain of discourse.
>neglecting that little factette and invalidating the proof ...
>
>Whenever
>iii aaa jjj .
>then
>jjj [rdf:type] yyy .
>hence
>jjj [rdf:type] zzz .
>
>hence
>
>aaa [rdfs:range] zzz .
>
>==
>
>I smell danger.
>

Relax. Even if it were RDF valid to infer from 'whenever xxx rdf:type 
aaa, xxx rdf:type bbb' to aaa rdfs:subClassOf bbb, which it isn't, 
that still wouldn't get you the rdfs:range inference because a 
subclass or superclass of a range needn't itself be a range.

Pat



-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes

Received on Friday, 28 June 2002 02:07:16 UTC