Finite-size scaling analysis of a nonequilibrium phase transition in the naming game model

We realize an extensive numerical study of the naming game model with a noise term which accounts for perturbations. This model displays a nonequilibrium phase transition between an absorbing ordered consensus state, which occurs for small noise, and a disordered phase with fragmented clusters characterized by heterogeneous memories, which emerges at strong noise levels. The nature of the phase transition is studied by means of a finite-size scaling analysis of the moments. We observe a scaling behavior typical of a discontinuous transition and we are able to estimate the thermodynamic limit. The scaling behavior of the clusters size seems also compatible with this kind of transition.