Not entirely perfect.  I noticed a bad typo and want to make sure I'm
not confusing anyone:

<In scenario A; the output would be triple 1 and 4. >

Should be

In scenario A; the output would be triple 1 and 3.

-- Scott

On Nov 10, 12:10 pm, PaulZH <[email protected]> wrote:
> Scott, a perfect explanation.
>
> Paul
>
> On Nov 10, 6:16 pm, Scott Henninger <[email protected]>
> wrote:
>
> > Paul;  First, it is the case that the asserted + inferred triples is
> > the result of running the ApplyOWLIM module.  However, when you do
> > that, the rdfs:subClassOf inferences will be made.  Your filter will
> > then filter out triples you hadn't before.
>
> > For example, let's say you start with:
> >   1.  :Matriarch rdfs:subClassOf :Person
> >   2.  :Person rdfs:subClassOf owl:Thing
> >   3.  :MyWife a :Matriarch
> >   4.  :Me a :Person
>
> > A variant of your filter is:
>
> > CONSTRUCT {
> >     ?a a ?type .
> >     ?a ?b ?c .}
>
> > WHERE {
> >     ?a a ?type .
> >     ?a ?b ?c .
> >     FILTER (?type != :Matriarch) .
>
> > }
>
> > In scenario A; the output would be triple 1 and 4.
>
> > If you then run OWLIM, you add the triples:
> >   5. :Matriarch rdfs:subClassOf owl:Thing
> >   6. :MyWife a :Person
>
> > The filter will now catch all of the instances, :MyWife in this case,
> > that have been inferred to be members of classed other
> > than :Matriarch.  Hence there are no results to pass on to the next
> > module.
>
> > -- Scott
>
> > On Nov 10, 10:12 am, PaulZH <[email protected]> wrote:
>
> > > Scott,
>
> > > The question is what is passed to the next module.
> > > Scott; OWLIM with default settings.
>
> > > In scenario A: the asserted triples
> > > In scenario B: the asserted triples + the inferred triples OR only the
> > > inferred triples.
> > > According to the what I see in the SPARQLMotion result pane it is
> > > asserted + inferred.
>
> > > So I would expect that if I have the same filter I at least get the
> > > same leftovers as in scenario A.
>
> > > Paul
>
> > > On Nov 10, 4:38 pm, Scott Henninger <[email protected]>
> > > wrote:
>
> > > > Paul; It all depends on what inferences are being made by OWLIM, hence
> > > > questions on what is in the model, what OWLIM ruleset is used (I'd
> > > > assume owl-max, the default) have to be asked.
>
> > > > -- Scott
>
> > > > On Nov 10, 6:44 am, PaulZH <[email protected]> wrote:
>
> > > > > I have following flow:
> > > > > a) import from remote sesame ->SPARQLMotion results as expected
> > > > > b) filterByConstruct with
> > > > > CONSTRUCT {
> > > > >     ?a a ?type .
> > > > >     ?a ?b ?c .}
>
> > > > > WHERE {
> > > > >     ?a a ?type .
> > > > >     ?a ?b ?c .
> > > > >     FILTER (?type != ns:ClassA) .}
>
> > > > > ->SPARQLMotion results as expected
>
> > > > > Now
> > > > > a) idem import from remote sesame
> > > > > b) ApplyOWLIM
> > > > > c) idem filterByConstruct with same query
>
> > > > > I get only a few triples with predicate rdf:first and rdf:rest.
>
> > > > > I would have expected the same result as previously.
>
> > > > > What am I missing?
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"TopBraid Composer Users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/topbraid-composer-users?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to