W3C home > Mailing lists > Public > www-webont-wg@w3.org > January 2004

Re: [jena-dev] Re: OWL subLanguage Determination

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Fri, 30 Jan 2004 09:29:27 +0000
Message-ID: <401A23F7.4070601@hplb.hpl.hp.com>
To: jena-dev@yahoogroups.com
Cc: www-webont-wg@w3.org


Cc-ing webont
Note: "Jeremy's jena.owlsyntax" is an "OWL Syntax Checker"

Kaile said:
 >   Managed to reduce simple query time from few minutes to .3 seconds
 > and without having to give the JVM any more memory too :))). Jeremy's
 > jena.owlsyntax really helped in tracing a lot of the ontology
 > definition problems.  Ensuring the sub-language being Lite improved
 > the merger and query time without having to make any code changes.
 >

(full context below)


I found this very surprising - you're using a rulebase reasoner that has no 
particular preference or desire for OWL Lite input (well not in the strict 
sense - I believe we have better rules for the Lite vocab), and you report 
that even so, the OWL Lite definition is a useful guide to gaining better 
performance.

I didn't expect that.

Jeremy




Context:
========
kailesmith wrote:

> Hello Ian,
> 
> --- In jena-dev@yahoogroups.com, "Dickinson, Ian J" > 
> There are three possibilities I can think of:
> 
>>i) you have an awful lot of data.  To test this, reduce the size of your
>>input set and measure the effects
>>
>>ii) you have found a pathological case in interactions between
>>
> certain OWL
> 
>>constructs and the Jena reasoner.  To test this, simplify your ontology
>>until the performance comes down to a reasonable level, then see
>>
> what you
> 
>>just took out and post it here :-)
>>
>>iii) you are setting your models up in a non-conventional way, and
>>
> that is
> 
>>exposing a bug, or at least an infelicity, in the Jena code.  You
>>
> can test
> 
>>this by posting your code here.  You said in your introduction that
>>
> you had
> 
>>"interesting times" trying to set up a model and perform a query. That
>>suggests to me that you are doing something out of the ordinary, because
>>setting up a simple test model and query is pretty straightforward
>>
> and not
> 
>>interesting in that sense!
>>
> 
>   Managed to reduce simple query time from few minutes to .3 seconds
> and without having to give the JVM any more memory too :))). Jeremy's
> jena.owlsyntax really helped in tracing a lot of the ontology
> definition problems.  Ensuring the sub-language being Lite improved
> the merger and query time without having to make any code changes. 
> 
> Almost all the files are now Lite, except for two. One because I am
> using unionOf (cannot avoid it) and the other is trying to put
> restriction on sequence membership (I have started another thread on
> that).
> 
> regards
> 
> 
> 
>  
> 
> Yahoo! Groups Links
> 
> To visit your group on the web, go to:
>  http://groups.yahoo.com/group/jena-dev/
> 
> To unsubscribe from this group, send an email to:
>  jena-dev-unsubscribe@yahoogroups.com
> 
> Your use of Yahoo! Groups is subject to:
>  http://docs.yahoo.com/info/terms/ 
> 
> 
> 
Received on Friday, 30 January 2004 04:32:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:03 GMT