Re: AW: Literals representing people?

On 2003-10-01 18:41, "ext LYNN,JAMES (HP-USA,ex1)" <james.lynn@hp.com>
wrote:

> So just to make sure I'm not missing anything, 1 has a node named _:x which
> is a blank node in 3

Both 1 and 3 have the blank node, and are identical expressions
of the same graph. In 1, a local identifier is used as NTriples
doesn't provide for the contracted form expressed in 3. But they
represent the same identical RDF graph.

> and is totally omitted in 2.

Right. 2 expresses a different graph than 1/3.

> But for these
> differences, they would all be equivalent.
> 
> 1    <#me> ex:myFriend _:x .
> _:x ex:emailAddress <mailto:somebody@example.com> .
> 
> 
> 2    <#me> ex:someFriendsEmail <mailto:somebody@example.com> .
> 
> 
> 3    <#me> ex:myFriend [ ex:emailAddress  <mailto:somebody@example.com> ]
> 
 
They wouldn't be explicitly equivalent. But given a pair of inference
rules relating the two forms, one could entail either of the two graphs
from the other. I.e.

IF
   ?s ex:myFriend ?o .
   ?o ex:emailAddress ?a .
THEN
   ?s ex:someFriendsEmail ?a .

IF
   ?s ex:someFriendsEmail ?a .
THEN
   ?s ex:myFriend ?o .
   ?o ex:emailAddress ?a .

Cheers,

Patrick