From: Brian McBride <bwm@hplb.hpl.hp.com>

Date: Tue, 24 Jul 2001 17:38:20 +0100

Message-ID: <3B5DA47C.2206DA1F@hplb.hpl.hp.com>

To: rdf core <w3c-rdfcore-wg@w3.org>

Date: Tue, 24 Jul 2001 17:38:20 +0100

Message-ID: <3B5DA47C.2206DA1F@hplb.hpl.hp.com>

To: rdf core <w3c-rdfcore-wg@w3.org>

I've done an updated version of the strawman model theory based on the feedback so far: Let U be the set of absolute URI References (as defined by RFC 2396). //ignore lang and namespaces for now Let S be the set of of ISO 10646 strings An interpretation I of STATEMENTS, a set of statements {s1, s2, ... ,sn} where each si is an n-triple statement, consists of: A set R of thingies A subset P of thingies which corresponds to Properties A mapping IN : U -> R A mapping IEXT : P -> powerset(R x R) // R cross R A mapping IS : S -> R Let STATEMENTS be a set {s1, s2, ..., sn} of n-triple statements Let A be the set of names of the form _:name in STATEMENTS STATEMENTS is true in I if and only if there is a substitution mapping SUBST from the members of A to R such that each member of STATEMENTS is true in I under that substition. <s> <p> <o> . is true in I under substitution SUBST if and only if: s, p, o are members of U, IN(p) is a member of P (IN(s), IN(o)) is a member of IEXT(IN(p)) <s> <p> "string" . is true in I under substitution SUBST if and only if: s and p are members of U, IN(p) is a member of P and string is a member of S (IN(s), IS(string)) is a member of IEXT(IN(p)) _:name <p> <o> is true in I under substitution SUBST if and only if: p and o are members of U, IN(p) is a member of P and (SUBST(_:name), IN(O)) is a member of IEXT(IN(p)) <s> _:name <o> is true in I under substitution SUBST if and only if: s and o are members of U, SUBST(_:name) is a member of P and (IN(s), IN(o)) is a member of IEXT(SUBST(_:name)) and similarly for anon nodes in other positions in the n-triple statement. No doubt there is a more elegant way of expressing this without lots of repetition. BrianReceived on Tuesday, 24 July 2001 12:40:55 EDT

*
This archive was generated by hypermail pre-2.1.9
: Wednesday, 3 September 2003 09:38:16 EDT
*