Re: Reification quoting in RDF/N3 was: A note comparing Conceptu al Graphs and RDF/Semantic Web

> IMHO RDF as-is lacks precision in this area; but I stick by my story that
> this is a problem for the Web at large. One that it would be nice to see
> folk on this list have a crack at...

This seems to relate to a discussion started by Pat Hayes last year on the use of
proper names in DAML and RDF.
  http://lists.w3.org/Archives/Public/www-rdf-logic/2000Oct/0112.html

To explain ... it seems more appropriate to me to use a proper name in the
assertion of what John believes. In effect saying "John believes that the person
he knows as 'Mary' wants to marry the person he knows as 'Fred'". Separate
assertions may be used to relate the (context-qualified) proper name to some
identifying information like a mailbox. Especially if it is actually 'Dan' who
believes he knows the mailbox for the 'Fred' that 'John' is talking about.

To encode the proper names like 'Mary' in RDF one could just use literals as (I
think) Bill suggests, or local uris as Sean suggests or a shared proper-name
namespace same as I suggested in the original discussion:
  http://lists.w3.org/Archives/Public/www-rdf-logic/2000Nov/0001.html

That would lead to something like:
 @prefix pn : <urn:daml-proper-name#>
  pn:John :believes {pn:Mary foaf:livesWith pn:Fred}.
  pn:Dan :believes {pn:Fred  foaf:mbox "mailto:fred@example.com" } .

This probably works better for more-nearly-unique names like 'Boston' but a
judicious use of anoymous nodes to associate context with the proper names could
make even this approach workable. So that the thing that 'Dan' believes becomes:
  {[ :knownAs pn:Fred :knownBy pn:John]  foaf:mbox "mailto:fred@example.com" } .

In turn this pattern of refering to person-A explicitly via the name they are
known by to person-B is a key concept in the SDSI/SPKI approach to security, see
http://theory.lcs.mit.edu/~rivest/sdsi11.html . So SDSI would refer to the "the
person known as Fred by John" using  (ref: John Fred). In the SDSI case the chains
of local references eventually bottom out at private/public key pairs rather than
mailboxes.

Dave

P.S. Apologies for my informal mis-use of the term "proper name" here but I
suspect I'd need to know much more than I do about signs and semiology to find
better terms.

Received on Thursday, 18 January 2001 13:10:24 UTC