[Attempto] RuleML question

Kaarel Kaljurand kaljurand at gmail.com
Fri Feb 6 15:33:25 CET 2009


Hi,

On Thu, Feb 5, 2009 at 6:11 PM, Joshua TAYLOR <tayloj at cs.rpi.edu> wrote:
>
> I notice that in the XML Schema description of the APE results [1]
> that the apeResult element may contain a ruleml element. However, in
> the specification of the there is no mention of ruleml. But, if I give
> the "solo" parameter a value of "ruleml", or provide a parameter
> "cruleml" with value "on", I get what I presume is valid RuleML [2].
> (I'm actually not familiar with RuleML myself.)
>
> Is this just an elision in the APE specification that should, perhaps,
> be updated? Something that is on its way out (and so removed from the
> specification, but not from the XML Schema)? This isn't a feature that
> I'm depending on, or even looking for—as I mentioned, I'm not familiar
> with RuleML—but just something I noticed.

"ruleml" is an undocumented parameter. :)

We implemented the RuleML support about 2 years ago but never really tested
if valid RuleML is generated in all cases, and never extended it to cover the
new language constructs that were introduced in ACE 5, 5.5, and 6.
So we didn't announce it either. The source code however is available,
it's in ape/utils/drs_to_ruleml.pl (in ape-6.0-081023.zip). This module
is really short and simple. If somebody wants to improve it and extend
it, then we
welcome patches. Otherwise it's not on our agenda to do anything with
this module
(apart from updating it if the DRS format changes).

Btw, the RuleML support was inspired by TRANSLATOR
by David Hirtle (see: http://www.ruleml.org/translator/). TRANSLATOR
accesses the
APE webservice, parses the DRS, and maps it into RuleML. All of this
is implemented
in Java. Our implementation is in Prolog.

--
kaarel


More information about the attempto mailing list