swrlb:subtract
Andy
swrlb:subtract (or any other builtin with more than two arguments). I
understand that you don't support inferencing of rules with builtins,
but I can at least use buildins like "(?x swrlb:greaterThan ?y)" and
have them pass the checker so that I can save the rule, which suggests
that I should be able to at least create a rule with swrlb:subtract?
Andy
Holger Knublauch
we only support the Jena rules language, not any SWRL built-ins. If you
can edit some SWRL built-ins then this is by coincidence. For things
such as subtraction, you could use the Jena rules language
http://jena.sourceforge.net/inference/#rules
Holger
Jody
I understand that you are using Jena and that the SWRL built-ins will
not be processed by the rule engine. But I am in a similiar situation
as Andy. I need to write rules that will be run in a separate system
(using JESS). We have our own built-ins for creating URI's and such.
I need to write the rules using these built-ins serialized SWRL/RDF or
ruleML. Currently I'm trying to push the use of Composer for ontology
management and editing. One of my internal selling points was that I
could write the SWRL rules in the same work environment.
Composer allows me to write SWRL rules and validates a certain
syntax. I'd expect that I could also write a rule that included a
swrl built-in and have the rule saved in rdf. The sytax could be the
same as Jena's and the syntax could be validated based on it being of
type swrl:Builtin and the number of arguments from the swrlb:args/
swrlb:minArgs/swrlb:maxArgs properties. I had assumed that this was
already implemented.
Jody
On Dec 18 2007, 2:17 pm, Holger Knublauch <hol...@topquadrant.com>
wrote:
>
> - Show quoted text -
Holger Knublauch
I see your point and I agree this is desirable. Our stand point is that
we would need a strong user pull to implement more features in this
direction. We need to balance all these requirements (and their
development and maintenance time) with our own resources in a relatively
small team. Having said this, we are hiring more people as we speak.
The current implementation uses the Jena rule parser to validate rules.
Then the Jena rule is translated into RDF SWRL representation if the
user is in SWRL mode. We would need to hook into the Jena rules parser
to allow custom built-ins or the built-in swrl built-ins. All is
possible, but a matter of priorities. Your opinion is valued here and
will likely have an impact on the future planning.
Holger
Jody
assume that in the previous post "SWRL mode" simply meant that the
SWRL ontology is imported / instantiated and that I haven't missed
some other functionality.
Jody
Holger Knublauch
Jody wrote:
> I appreciate your feedback and that you need to prioritize. Can I
> assume that in the previous post "SWRL mode" simply meant that the
> SWRL ontology is imported / instantiated and that I haven't missed
> some other functionality.
Yes, with SWRL mode I just meant rules that are expressed in SWRL
syntaxt. You can in theory use Jena rules and SWRL rules side by side
in the same file, but in both cases Jena will be used for execution.
Holger