Analysis of naming game based collective behavior with biased assimilation over adaptive networks

Abstract

The dynamics of two-word naming game incorporating the influence of biased assimilation is investigated in this paper. Firstly an extended naming game with biased assimilation (NGBA) is proposed. The hearer in NGBA accepts the received information in a biased manner, where he will refuse to accept the conveyed word with a predefined probability, if it is different from his own current memory. Secondly, the adaptive network is formulated, where the existing links may be reconnected so that the individual can influence other individuals over the population as much as possible. Numerical simulation results show that the population in NGBA can eventually reach consensus; the larger strength of biased assimilation on both words, the slower convergence speed, while larger strength of biased assimilation on only one word can slightly accelerate the convergence; larger population size can make the rate of convergence slower to a large extent when it increases from relatively small size, but such effect becomes minor when the population size is large; the behavior of adaptively reconnecting the existing links can greatly accelerate the rate of convergence.