Use of DAML+OIL/SkiCal and iCalendar in DAML+OIL

Mon Oct 7 02:43:24 PDT 2002

Greg,

Thanks for pitching in here. I think Alan's comment that 'it is very
important to know, concretely, what causes problems to others in other
fields' is exactly right. 

I guess what applications people (which I mostly fit into) are looking
for is some idea of the types of problem domain that an ontological
approach will and will not fit. And the ontology people are looking to
iron out the problems that do arise.

It seems to me that the SkiCal experiment was akin to problems in the
business arena. In the past I've spent a long time trying to define the
structure of a problem, mapping the right classes and structures that,
together, will make it easiest to instantiate business objects and
processes and then building an application around them. The uncertainty
here was in getting the right class to map a set of objects. Lots of
mappings could work; the trick was choosing the best one. 

Given that you'd already done this to get the application working, the
ontological work was of little extra benefit. And, even if starting from
scratch, the effort involved was greater than the original modelling
effort. No?

In the scientific domain, such as with astronomical VOs, it is
different. While we still have the issue of determining the right way to
map concepts as classes, whether using a data modelling or ontological
approach, there is an additional problem with creating the instances.
Three people may analyse the same data for a given point in the sky: one
will say the object is a brown dwarf, one will say it is a 'green fairy'
and the third may say that the readings are simply errors in the
instrument. All three 'statements' must be possible about the same set
of data points. 

The additional benefit we're looking for from an ontological approach is
the use of an inference engine to tell us when one set of statements is
the same or different to another set. Whether that can then be put to
practical use is another question.

Or perhaps what we're looking for is a way in which the set of data
points is only ever represented by these 'statements'. The 'object' is
never instantiated as an instance of one or other classes - bit like
quantum superposition except that resolution is never absolute :) I
wonder if this will affect the way we design our VO more fundamentally
than I though - worth thinking about.

<humming sound as Tony disappears into another reverie...>

Cheers,
Tony. 

> -----Original Message-----
> From: Greg FitzPatrick [mailto:greg at skical.org] 
> Sent: 07 October 2002 09:32
> To: rector at man.ac.uk; tol at star.le.ac.uk
> Cc: semantics at us-vo.org; danbri at w3.org; 
> libby.miller at bristol.ac.uk; Par Lannero
> Subject: RE: Use of DAML+OIL/SkiCal and iCalendar in DAML+OIL
> 
> 
> Alan and Tony
> 
> It has been a while since I worked on this and my memory 
> isn't all that great.
> 
> I remember that I was quite enthusiastic about the annotated 
> walkthru created to exemplify  DAML+OIL.  It seemed to me 
> that we could just rewrite the walkthru using SkiCal instead 
> of the existing examples and I started out with that 
> intention, replacing animals with events and so on.
> 
> The further I got into this experiment, the more apparent it 
> became to me how artificial the exemplifying ontology of the 
> walkthru was.  Yes, of course it was just a bunch of made up 
> examples, but as I progressed it became increasingly 
> difficult to discern instance from schema.  There seemed to 
> be a slightly tautological aspect to the entire walkthru 
> which I had missed prior to my attempt to reuse it.  
> 
> Obviously there is a danger in picking fictional examples out 
> of your head (populating an idea) to exemplify a structure 
> since those examples are always going to fit.   
> 
> I am sure all these problems are old hat for others and I am 
> surely the least qualified to criticize DAML+OIL. After the 
> fact I would say that SkiCal just wasn't "hierarchal" enough 
> to make sense in DAML+OIL and I am not quite sure what would 
> be, but Alan has asked for a response, maybe we can talk a 
> bit about this on Wednesday and I can brush up on the work 
> and be more concrete, otherwise I realize that this too is an 
> "argument with out evidence".
>             
> Greg
> 
> P.S.
>   
> I have jokingly said to Libby and Dan that I think all schema 
> examples should be about aliens and written in at least 
> Russian, or some other language or topic that those involved 
> in the schema work do not understand.  I am sure the 
> "self-evident" nature of the IPO example invariably used in 
> schema walkthroughs (though not in the DAML+OIL instance) 
> have caused considerable damage.
> 
>        
> 
> 
> 
> -----Original Message-----
> From: Alan Rector [mailto:rector at cs.man.ac.uk]
> Sent: Sunday, October 06, 2002 12:19 PM
> To: tol at star.le.ac.uk
> Cc: semantics at us-vo.org; greg at skical.org; danbri at w3.org; 
> libby.miller at bristol.ac.uk
> Subject: Re: Use of DAML+OIL/SkiCal and iCalendar in DAML+OIL
> 
> 
> Tony - Dan, Libby and Greg
> 
> Thanks to Tony for bringing the paper to our attention.
> 
> A question for the authors...
> 
> Most of the issues concern namespaces and equivalence which I 
> shall leave to Sean.
> 
> Towards the end of the paper you have an almost throw away 
> comment that
> 
>     "Another significant problem was the awkward syntax and 
> limited expressivity..."
> 
> Leaving aside the awkward syntax, I would be curious to know 
> specific examples of the  problems with expressivity in your 
> concrete task.  The reference to Pat Hayes isn't helpful 
> because I know his theoretical position.  The question is, 
> "What happens in practice?"
> 
> I know which things have caused problems in  building 
> biomedical ontologies.  I think it is very important to know, 
> concretely, what causes  problems to others in other fields. 
> Otherwise we are arguing without evidence.
> 
> Thanks
> 
> Alan
> 
> --
> Alan L Rector
> Professor of Medical Informatics
> Department of Computer Science
> University of Manchester
> Manchester M13 9PL, UK
> TEL: +44-161-275-6188/6239/7183
> FAX: +44-161-275-6204
> email: rector at cs.man.ac.uk
> web: www.cs.man.ac.uk/mig
>         www.opengalen.org
> 
> 
> 
> 
> Tony Linde wrote:
> 
> > Ran across this paper at:
> >   http://www.ilrt.bristol.ac.uk/discovery/2002/03/skical-daml/
> >
> > It deals with applying DAML+OIL approach to an existing application 
> > (SkiCal). I'd be interested to get Sean's comments on how they 
> > approached the problem and their conclusions about DAML+OIL.
> >
> > (I'll be glad when OWL takes over - it's easier to type!)
> >
> > Cheers,
> > Tony.
> >
> > __
> > Tony Linde                       Phone:  +44 (0)116 223 1292
> > AstroGrid Project Manager        Fax:    +44 (0)116 252 3311
> > Dept of Physics & Astronomy      Mobile: +44 (0)7753 603356
> > University of Leicester          Email:  tol at star.le.ac.uk
> > Leicester, UK   LE1 7RH          Web:    http://www.astrogrid.org
> >
> 
> 
> 
> 