[redland-dev] Can Redland process transitive properties when running Sparql queries?

[Slava Kravchenko]

On Monday 27 Aug 2012 12:00:01 redland-dev-request at lists.librdf.org wrote:
>    1. Can Redland process transitive properties when	running Sparql
>       queries? (Seref Arikan)

[...]

> Greetings,
> I need to process rdf in postgresql functions, and my plan is to use Python
> wrappers for Redland to process turtle synax rdf, with some Sparql queries.
> 
> Jena helps me on the client side by making use of transitive properties, so
> when in my ontology a relation among nodes such as this one exists:
> 
> A contains B contains C
> 
> A Sparql query run by Jena for ?Node :contains ?childNode
> returns both A B and A C pairs, due to transitive nature of contains
> property. Protege can not make use of this information for example. so the
> same Sparql query returns only A contains B when run from Protege.
> 
> If I use Redland, would it be able to act in the same way as Jena? Your
> response would help me skip installation time for the whole stack to find
> the answers on my own and it would be appreciated a lot :)

I believe the task of producing extra triples based on property transitivity 
(and similar cases) is the one that Redland "delegates" to the underlying 
storages. So, if your storage has reasoning capabilities (e.g. Virtuoso, 
although I couldn't make it work here), it will send the whole bunch, both 
real and "computed" RDF triples to Redland.
So, ultimately, Redland cannot be compared to Jena - it's not a "semantic 
framework", but rather, a simple "interface" to a semantic storage.

-- 
Best regards,

Slava Kravchenko