[
Lists Home |
Date Index |
Thread Index
]
John Cowan wrote:
> Tim Bray scripsit:
>
> > Questions: (a) can software detect the difference?
>
> Not as RDF stands, no.
As I've said, RDF is itself quite capable of asserting these differences.
>
> > (b) does the difference have any effect on the behavior of the system?
>
> Definitely, since "the system" includes human beings and other
> inference-drawing machines. Identifying John Cowan with John Cowan's
> mailbox or John Cowan's homepage leads to paradoxes, since the same
> properties have different values for these things: I began in 1958,
> my mailbox in about 1993, and my homepage not until 1998. A pretty
> drastic distinction!
>
Pat makes an excellent point that the difference between 1) and 2) DOES have
an effect on inferencing engines ... these are existing software products,
e.g www.networkinference.com, FACT, Protoge, OilEd. HP has a substantial
group working in the U.K. on this sort of thing.
The sorts of facts you discuss above are exactly the sorts of facts that can
be easily represented in RDF and from which inferences can be drawn using
existing OWL or DAML+OIL software.
Now of course if you define "the system" as the current human readable Web,
and exclude the human readers, then no, that narrow definition of the system
cannot make the distinction. But as John says, if we include the human
agents which currently use the human readable Web, or if we include the
machine agents which operate on the machine readable Web (agreed this second
web is much much smaller) then Yes, the system notices the difference.
I am coming around to position (1).
Jonathan
|
