Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[protege-owl] SWRL execution in protege 3.4

Martin O'Connor martin.oconnor at stanford.edu
Mon Mar 30 09:15:47 PDT 2009


In 3.4, SWRL rules are not sent to Pellet. Until recently, Pellet had 
only partial SWRL support. Now it supports nearly all of the the 
language defined in the SWRL Submission. I intend to update the SWRLTab 
to send SWRL rules to Pellet in the next month.

It is also worth noting that the SWRLTab has an array of built-in 
libraries that will not be supported by Pellet (e.g., SQWRL [1]) and 
will only work with the Jess back end.

I hope to have an additional completely open source rule engine back end 
for the SWRLTab later this year.

Martin

[1] http://protege.cim3.net/cgi-bin/wiki.pl?SQWRL

Pallinger Péter wrote:
> Thank you for the fast reply!
>
> Michael Lodemann írta:
>   
>>> Class_1(?x) -> Class_2(?x)
>>> I could create the rule, but I failed to execute it (of course, I
>>> created both classes,
>>> and added an individual belonging to Class_1). Is there a special menu
>>> item to do this?
>>> Is it possible at all?
>>>       
>> What kind of error shows up? 
>>     
> There is no error message, it just doesn't do anything.
> I would think that because pellet is built-in, and pellet supports SWRL (that very
> simple part that I have used for sure), it should automatically infer classification
> based on the rules I give.
>
>   
>> Did you install the Jess rule engine within
>> the SWRLTab? Perhaps this link helps you:
>> http://protege.cim3.net/cgi-bin/wiki.pl?SWRLJessTab
>>     
> I have tried that now.
> It is better than nothing, but I have some problems with it:
>  1. The result of Jess'es inference appear as _asserted_ individuals. Although this may
> be acceptable for small ontologies for testing out a small number of rules, it is not
> acceptable for larger ontologies where one may not know all instances and classes by
> heart and may not remember all previous assertions.
>  2. There is an import-run-export procedure involved which is somewhat cumbersome.
>  3. Jess is a proprietary solution whereas the already included open-source pellet is
> available to do most SWRL inference. Why not use that?
>
>   
>>> In protégé 4 (Version 4.0.109), both pellet 2 and pellet 1.5 manage to
>>> infer that the
>>> given individual indeed belongs to Class_2, but I cannot edit SWRL
>>> rules in protégé 4.
>>> Is that really so, or just I am clumsy?
>>>       
>> As far as I know protege 4 doesn't support SWRL rules yet. You must use
>> protege 3.4 in order to execute SWRL rules.
>>     
> On the contrary! If you would have read the passage above, you would know that I
> actually succeeded in _executing_ SWRL rules in Protégé 4.0 using both pellet 1.5 and
> 2.0 (not at the same time, of course :) ). The only thing I miss from Protégé 4.0 is an
> SWRL editor (preferably using SWRL abstract syntax).
>
> I am still somewhat stuck, as SWRLJessTab is not a really good solution for SWRL
> inference in Protégé 3.4.
>
> Thanks in advance for any help.
>
> 	PP
>
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> protege-owl mailing list
> protege-owl at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03 
>   




More information about the protege-owl mailing list