Re: Semantics of Preconditions and Effects

On Wed, 27 Aug 2003, Monika Solanki wrote:

> Hi Sheila,
>
> Thanks for your Reply.
>
> Sheila McIlraith wrote:
>
> >Hi Monika,
> >
> >As you know, DAML-S enables the modeler to associate the following
> >properties with a process
> >- input
> >- (conditional) output
> >- precondition
> >- (conditional) effect
> >
> >"inputs" and "outputs" should be obvious.  A "precondition" describes some
> >state of the world that must be true in order for the service to be
> >invoked.  "(condiitonal) effect" describes the side effect of the program.
> >E.g., if executing the amazon.com service has the side effect that a
> >book is shipped to the address provided, then this is encoded as a
> >"(conditional) effect".  Note it is conditional if the effect occuring
> >depends on some condition.
> >
> >As we argued in the paper you referenced below, "inputs" and "outputs"
> >can be regarded as "knowledge preconditions" and "knowledge effects".
> >That is, they pertain to the agent (executors) state of knowledge.
> >If ISBN is the "input" to amazon, it is identical to say that the agent
> >must *Know* the ISBN number;  this is a knowledge precondition of the
> >process.  Likewise, if the "output" is the price of the book,
> >then it follows that the knowledge effect of the process is that the
> >agent *Knows* the price of the book.
> >
> >So, to answer your question, the agent can have knowledge effects, but
> >they are encoded as and "output" of a process.  That is the way
> >the agent's state of knowledge is updated.  Note also that the agent
> >*Knows* all the effects of the process it is invoking, so it follows
> >that any "effect" is also a "knowledge effect".
> >
> So does it actually mean that these knowledge preconditions and effects
> are ineffect  manifestations of "input" and "output" and different from
> the actual preconditions and effects that we usually need for a DAML-S
> specification?



Yes.  that is correct.


> I mean, can we actually use these knowledge preconditions
> and effects to specify the preconditions and effects of the service?


You can use the DAML-S/OWL-S "input" and "output" properties to
specify knowledge preconditions and knowledge effects.  This is the
correct interpretation.

By the way, I realized, this topic may have been discussed more in

McIlraith, S., Son, T.C. and Zeng, H. ``Semantic Web Services" , IEEE
Intelligent Systems. Special Issue on the Semantic Web. 16(2):46--53,
March/April, 2001. Copyright IEEE, 2001.


- Sheila McIlraith


> >Sheila McIlraith
> >
>
> --
> **>><<**>><<**>><<**>><<**>><<**>><<**>><<**
> Monika Solanki
> Software Technology Research Laboratory(STRL)
> De Montfort University
> Hawthorn building, H00.18
> The Gateway
> Leicester LE1 9BH, UK
>
> phone: +44 (0)116 250 6170 intern: 6170
> email: monika@dmu.ac.uk
> web: http://www.cse.dmu.ac.uk/~monika
> **>><<**>><<**>><<**>><<**>><<**>><<**>><<**
>
>
>

==============================================================================

Sheila McIlraith, PhD                 Phone: 650-723-7932
Senior Research Scientist             Fax:  650-725-5850
Knowledge Systems Lab
Department of Computer Science
Gates Sciences Building, 2A-248       http://www.ksl.stanford.edu/people/sam
Stanford University                   E-mail: sam-at-ksl-dot-stanford-dot-edu
Stanford, CA 94305-9020

Received on Wednesday, 27 August 2003 13:38:09 UTC