[Attempto] loading, saving, and clearing ACE Wiki ontologies? Controlled languages built on other formalisms?

[Norbert E. Fuchs]

On 5 Dec 2009, at 00:59, Murray, William R wrote:

> On another academic thought of interest I was just wondering to what extent ACE had been
> translated into formalisms other first-order logic and description logic? It seems like there is
> ample room for a whole family of alternate formalisms from constraint description languages
> to Markov Logic Networks to Bayesian Networks, etc, etc. And seems like one could add a
> temporal reasoner for Allen’s Temporal Algebra. Lots of thesis material if you have enough
> grad students in the pipeline :)

Bill

Let me answer in three steps.

1. ACE contains constructs that are within first-order logic, and other constructs that go beyond what is usually considered standard first-order logic, for instance , 
negation as failure
sentence subordination that could be translated into first-order logic using possible world semantics
forms of modality (necessity, possibility) that could be translated into first-order logic using possible world semantics
forms of modality (recommendation, admissibility) that do not seem to fit into first-order logic
interrogative sentences
imperative sentences

2. The ACE parser APE generates for each ACE text a DRS that uses a syntax that we extended to accommodate all of the above constructs

3. Some parts of a DRS can be translated into standard language of first-order logic, other parts would need first-order logic with possible world semantics, while still other parts cannot be translated into first-order logic at all.

4. However, the restrictions (3) do not prevent you from translating the DRS into languages different from first-order logic. An example is Tobias Kuhn's AceRules that, for instance uses negation as failure. Another example is a cooperation with Yale University where we express clinical guidelines that use all the above forms of modality as a rule-based system.

Thus the way is wide open for your suggestions, though I am afraid that we do not have any grad students in the pipeline.

Regards.

   --- nef
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ifi.uzh.ch/pipermail/attempto/attachments/20091205/6c4c5f24/attachment.htm>