Distributed Artificial Intelligence
Research Program - Results (6)
GAg, a model of generic agent
Developing multi-agent systems suffer from the lack of generic agent. At
first, we tried to combine models of agents offered by other research groups
(Demazeau and Muller, Gasser, Hayes-Roth, Shoham, Sycarra, Maes, Woolridge
and Jennings, and others). We found that the models were quite different
and were incomplete. Thus, we used the different models to design a new
model of cognitive agent which we hope will be usefull to make some progress.
Our model GAg (Generic Agent),
allowed us to develop an agent language for creating agents, thus freeing
the user from the difficult task of engineering her own architecture.
Return
Last modified 8/4/02