[Attempto] algebraic expressions in ACE - continued

[Jean-Marc Vanel]

2010/6/14 Kaarel Kaljurand <kaljurand at gmail.com>:
...
> Maybe it could already now be a solution for you, i.e. forget about
changing
> owlfss_owlrdfxml.pl, and try to connect the output of ACE->OWL/SWRL
> (drs_to_owlswrl) directly to Thea's RDF/XML writer (if such a thing
exists).

I also think that it makes sense to reuse existing tools, and Thea looks
good, and I exchanged mails with Vangelis some mounths ago.

However, for the algebraic expressions, the worm in the fruit
drs_to_owlswrl, not only in owlfss_owlrdfxml.pl .
More precisely, I updated
ape/utils/owlswrl/implication_to_swrl.pl
so that I-variable/1 predicates are added recursively around DRS/SWRL
variables.

Now the OWL FSS looks good:
      Consequent(
         description(
            Class(owl:Thing)
            I-variable(6)
         )
         description(
            ObjectProperty(:has-surface)
            I-variable(1)
            I-variable(6)
         )
         builtIn(multiply, I-variable($VAR(6)), I-variable($VAR(2)),
I-variable($VAR(4)))
      )

I commited all my modifications here on Subversion (same directory
structure) :
http://deductions.svn.sourceforge.net/viewvc/deductions/ape/
<http://deductions.svn.sourceforge.net/viewvc/deductions/ape/>
This is work in progress; I get an SGML exception while writing the OWL/RDF
file.

I think I'll go on hacking the existing (and brushing my rusted Prolog),
while also paving the way with Vangelis for using his tools. I'm not sure at
all that his predicates for OWL are similar to ACE's, not to speak about his
rule representation.

In issue that is mentionned in ACE articles is that datatype properties are
hard to detect.
In the example mentioned before:
*If a n:rectangle R has a n:length L1 and R has a n:length L2 then R
v:has-surface S, and S = ( L1 * L2 ) .*

has-surface is wrongly taken as an ObjectProperty.
It could however be detected as a numeric datatype Property because of the
formula that produces a numeric result.

-- 
Jean-Marc Vanel
Consulting, services, training,
Rule-based programming, Semantic Web
http://jmvanel.free.fr/
+33 (0)6 89 16 29 52 -- +33 (0)1 39 55 58 16
( we rarely listen to voice messages, please send a mail instead )
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ifi.uzh.ch/pipermail/attempto/attachments/20100617/d17b1804/attachment.htm>