Re: Inference in daml

On June 17, Jim Hendler writes:
> At 11:04 PM +0100 6/17/01, Ian Horrocks wrote:
> >On June 17, Geoff Chappell writes:
> >>  Hi folks,
> >>
> >>  I've been working with expressing inference rules in daml and need 
> >>a little help/feedback.
> >>
> >>  It seems that rules with just the subject unbound can be expressed easily.
> >>
> >>  For example the rule:
> >>      type(X,animal)<-type(X,dog)
> >>  can be expressed as:
> >>      type(X,animal) or not(type(X,dog))
> >>  or in daml:
> >
> >It seems to me that all you are saying here is that dog is a subClassOf
> >animal. What is wrong with
> >
> ><daml:Class rdf:ID="dog">
> >   <rdfs:subClassOf rdf:resource="#animal"/>
> ></daml:Class>
> >
> >Am I missing something?
> >
> >Ian
> 
> Ian-
>   You and Jeff Heflin had a discussion at one point about what sorts 
> of SHOE [1] rules could and couldn't be expressed in DAML.  Did that 
> ever get written down?  Seems like it would be useful in helping 
> Geoff (who later wrote)

We didn't ever write it down. I'm not sure that we came to any
startling conclusion, but I will speak to Jeff and see if we think we
can come up with some notes that would help Geoff and others with
similar requirements.

> 
> At 4:42 PM -0400 6/17/01, Geoff Chappell wrote:
> >
> >Thanks for the response, but... I guess I need to be careful about my
> >(over)use of adverbs -- "ultimately"  I'm not trying to express anything
> >about dogs or animals necessarily but to translate inference rules of all
> >(or some) types into daml terms (if possible).  My example was a bad one
> >because there are so many ways without explicit inference to get the point
> >across (as you've demonstrated).
> 
> My recollection is that DAML can do many things, but there are many 
> rules one might wish to express that aren't easily done in DAML

This is certainly true.

> Have a good reference on this?

Not that I can think of. Comparing the expressive powers of different
logics is notoriously difficult.

Regards, Ian