Adaptive Automata Community Detection and Clustering: A generic methodology

We present in this paper a generic methodology based on genetic automata for modelling community detection. With Communities, we deal with dynamic organizations which are self-organized from two aspects, the spatial one and the functional one. We propose in this paper a general methodology which extends cellular based modelling like Schelling models to more sophisticated approaches based on agent behavior modelling. These agent behaviors are modelled by automata with multiplicities. These automata-based models allow to define powerful operators like genetic operators and behavioral semi-distance. Mixing these operators, we can propose a complex computing, dealing with spatial selforgnaization coupled with behavioral similarity for adaptive self-organized systems.