[Pellet-users] Accelerating Pellet Classification

[sjtirtha]

Hi,

I classify my ontology over pellet API. It takes around 3 seconds for a
simple ontology and it takes around 15 seconds for a complex ontology.
We're planning to use pellet in our business process, which should be able
to process 20000 requests in a hour. Processing the request includes
subsumption of pellet.
It seems to be not a realistic requirement. But still I want to accelerate
the reasoning under 1 second.

I see on my console, if I classify my ontology against pellet. That pellet
does some consistent check, that can be done on the build time.
For example, pellet checks every single class whether it is consistent or
not. In my opinion this can be done when we model the ontology. Can I
deactivate this function ?
I can understand that checking the combination of classes should be
processed in the runtime.
Or maybe can I annotate some classes that they don't need to be classified ?
So pellet just passes these classes in classification process.

Regards,

Steve
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.owldl.com/pipermail/pellet-users/attachments/20070627/06338e51/attachment.htm