Re: [namedgraphs] Reasoning with Named Graphs

[Graeme S. <ac...@gm...>]

On 9 February 2011 21:44, Marcus Cobden <li...@ma...> wrote:

> One of the things I wanted was for the inferred triples to be placed in
> different graphs, depending on which graphs the antecedents came from. Would
> that be possible?
>
>
Yes, this is straightforward. As an example, here's the Jess version of the
RDFS rule for inferring the type of a resource from a property domain:

  ;;; prp-dom: infer resource type from property domain

  (defrule prp-dom (declare (salience 100))
   (logan-quad
    (graph ?g)
    (subject ?p)
    (predicate "http://www.w3.org/2000/01/rdf-schema#domain")
    (object ?t)
   )
   (logan-quad
    (graph ?h)
    (subject ?s)
    (predicate ?p)
    (object ?o)
   )
   =>
   (assert
    (logan-quad
     (graph "urn://inferred")
     (subject ?s)
     (predicate "http://www.w3.org/1999/02/22-rdf-syntax-ns#type")
     (object ?t)
    )
   )
)


While in this instance the consequent asserts the entailed triple in the
"urn://inferred" graph, you can easily set this to be ?g, ?h (the graphs in
the anticedents), or define a function that selects somewhere else for it to
go.


>
> Btw: I spent an hour or so today adding generics to all the NG4J classes,
> if you're working with NG4J it might make things a bit easier ;) I'll be
> posting the diff tomorrow.
>
>
Sounds useful!

Cheers,
Graeme.