[Modeling] Modeling an Agent Class- register your opinion

Stephen Cranefield scranefield@infoscience.otago.ac.nz
Tue, 24 Jun 2003 09:45:51 +1200


Jim Odell wrote:
> OK, the tally is now:
> 
> --Caste - 1 vote
> --Agent Class or Agent Type - 5 votes
> --Agent - 1 vote
> --"The term agent class is 'misfortune'" - 1 vote
> --"No idea"  - 1 vote

I favour AgentType if we decide to use a metaclass that is a direct subclass
of Classifier (i.e. not a subclass of Class).  Otherwise, AgentClass would
be appropriate.

I also think that we should start developing a metamodel in which the agent
type metaclass *is* a new subclass of Classifier.  If we later discover that
it really can fit under Class then we can refactor the metamodel.

- Stephen