RE: timbl-03

> >2: the triples *are* redundant
> >    It is easy to automatically add them; or for that matter, automatically
> >remove them.

> Adding them is a valid entailment. Removing them is not.

The two entailments I had in mind were

_:a rdf:type rdf:List .
_:a rdf:first any .

=>

_:a rdf:first any .
(i.e. deleting the type triple)

and

_:a rdf:type rdf:List .
_:a rdf:rest any .

=>

_:a rdf:rest any .
(i.e. deleting the type triple)

In the presence of the domain and range constraints then the RHS graphs are 
semantically equivalent to the LHS graphs. I was not intending the unsound.

_:x rdf:type rdf:List .
=>
(empty)

where the RHS graph has fewer entailments than the LHS graph.

Jeremy

Received on Tuesday, 6 May 2003 16:15:59 UTC