[Modeling] Modeling an Agent Class- register your opinion

Joaquin Peņa joaquinp@us.es
Tue, 24 Jun 2003 10:34:58 +0200

Hi all:

Although, this TC is only for modelling i think we must take into
methodology issues to provide them the proper tools.

At analysis stage collaboration can be used to understand easily
the system making a "direct" mapping from requirements, to later,
assign roles in collaborations to agent-classes.

The key point is that for methodologies purpose collaborations
must be taken into account in this TC identifying if we need any

PS: I have misunderstanding a point: Does an agent class have a
set of roles it can perform defined explicitly? I think it should.


> On 6/23/03 4:33 AM, "Joaquin Peņa" indited:
> > That is our idea: Use role modelling techniques at analysis 
> stage and 
> > use design patterns at design that allows to implement roles 
> > decoupling coordinated behaviour (interactions with other 
> agents) and 
> > functionality (services, knowledge, abilities, ... Agents have).
> Am having some difficulty in a few areas here:
> 1) What do you mean by analysis vs design.  I define analysis 
> as the mapping from some perception to an 
> implementation-independent model of the perception (PIM); 
> design I define and the mapping from an 
> implementation-independent model to an 
> implementation-dependent model (PSM).
> 2) Given this, I am not sure what exactly you are putting 
> into the analysis (or PIM) level, and what you are forbidding 
> to go into that level -- because I do a lot a modeling at this level.
> 3) Similarly for the  PSM level: I am not sure what exactly 
> you are putting into the PSM level, and what you are 
> forbidding to go into that level.
> Lastly, am not sure that the Modeling TC should be addressing 
> methodology issue -- at least at this point.  These are 
> issues for the Methodology TC. That way, we can focus on what 
> we need to do to support agent-based modeling *for any 
> purpose* whether it be for analysis, design, or 
> implementation. That's why we created to different TCs.
> -Jim
> _______________________________________________
> Modeling mailing list
> Modeling@www.fipa.org
> http://fipa.org/mailman/listinfo/modeling