Moduleco
|
In this model, social network is formalized as a neural network with a small degree of connectivity (say in average 1,5%). After learning, some people will became the center of larger network (say in average 10%) as act as kind of "leader of opinion". as a result, we have an inhomogeneous process of diffusion, with classical S- shaped curve, but non classical increments, which are following a power law distribution.
From Plouraboué, Steyer, Zimmermann [1998] Deroïan [1999], Steyer, Zimmerman [2000], Deroïan, Steyer Zimmerman [2000]
for more information see (in French) :
http://www-eco.enst-bretagne.fr/~phan/complexe/fpasjbz.htm
Denis.Phan@enst-bretagne.fr ; Antoine Beugnard@enst-bretagne.fr