On Jun 2, 2010, at 12:25 AM, Schmitz, Jeffrey A wrote:

> On a related note, but more specific to Topbraid, I noticed that working on a 
> large model (about 2,000,000 triples), Topbraid seems to work a little 
> differently.  For example, when opened in Topbraid,  it doesn't seem to 
> perform the normal "default" inferencing it usually does automatically for 
> smaller models?   For example, class/subclass inferencing isn't performed, 
> nor do the instance counts exist.  Is the correct?

Hmm, not that I am aware of. I checked our source code and also didn't find any 
such size limitations. Computing the number of triples is already an expensive 
operation on some databases. There are some small differences in the behavior 
with database projects in general, no matter what size they have. One 
difference is that the system will switch to "direct instances" mode only and 
not show the number of instances of sub-classes. Also, the Instances view 
cannot be sorted by columns (the latter has been relaxed for TDB from 3.4 
onwards because TDB is fast enough for large number of small transactions).

>  I'm guessing the answer as to why this would be is that to try and do 
> otherwise would simply be too memory intensive.
> As I'm working with larger and larger models I'm starting to come to the 
> conclusion that if you have models of any size (say on the order of 1,000,000 
> triples) and you want to do any inferencing (even OWL_MEM) you will need a 64 
> bit machine.  Would you say this is a fair statement?

I have no experience with Jena inferencing (OWL_MEM etc) of that size, but my 
guess is that such things will work best if you are using a local database 
(TDB), and TDB is optimized to work on 64 bit machines.

Holger

-- 
You received this message because you are subscribed to the Google
Group "TopBraid Suite Users", the topics of which include TopBraid Composer,
TopBraid Live, TopBraid Ensemble, SPARQLMotion and SPIN.
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-users?hl=en

Reply via email to