Interactive particle-based simulation of sociophysics models

Abstract

Particle-based simulations are widespread in the field of computer graphics and are mainly used for simulating soft-body dynamics. Modeled objects are approximated through discrete points called particles, having various physical properties. Sociophysics has emerged alongside econophysics as a new important research domain, taking methodological inspiration from lattice model computer simulation of problems in two-dimensional condensed matter physics. This paper presents an extension of particle models with regard to their initial purpose and proposes a solution for simulating sociophysics models interactively using a particle-based visual approach. An existent agent-based “small world” model is mapped on a particle-based grid and its evolution in time is simulated on a high-performance graphics cluster in order to model technology adoption and consumer behavior. Several experimental scenarios validate the initial hypothesis that particle-based models can be extended beyond their original scope and evaluate the system's performance and scalability.