Abduladem,

you can use both CONSTRUCT and INSERT to infer new triples.

Both OWL and SPARQL have their advantages. OWL forms ontologies that you
can import etc., but often it produces too many (trivial) triples that you
don't care about, and performance can be poor. With SPARQL you can be more
precise with your patterns and execute faster.

As always, it depends on the use case. Given that few triplestores offer
inferencing and we don't want to reason over the whole dataset, we use
SPARQL to infer things like inverse properties, and then when the result is
loaded, we use OWL over the (small) result to infer subclasses and types
etc.

Hope it helps.

On Fri, Oct 6, 2017 at 4:09 PM, Abduladem Eljamel <
[email protected]> wrote:

>
> Hello,,
>
> My simple understanding of Inferencing is “inferring new facts
> fromexisting facts”. In that sense, I believe that there are some
> similarities of INSERT command in SPARQL.
>
> Can SPARQL be utilised in Inferencing?
>
> Are there any researches in investigating the cons and pros of usingSPARQL
> as reasoner?
>
> Thanks
> Abdul
>

Reply via email to