A Brownian Agent Model for Analyzing Changes in a Nation's Product Space Structure

Abstract

This paper proposes a Brownian agent model for simulating and analyzing changes in a nation's product space structure. A measurement of proximity has been employed to quantify a relationship between two products and proved to be useful in product space analysis. This study employs such proximity measurement, and estimates a continued structure transformation of a nation's product space through feedback between agent movements and network evolutions. Labor resources of an enterprise or a firm are regarded as Brownian agents; they move through different product spaces for higher economic rewards. The simulation results show that trade areas were self-organized through Brownian agent migration and cooperative production with a random initial distribution. Furthermore, we have verified the applicability and efficiency of the model in analyzing changes in Chinese product space structure with empirical data. Main contributions of this paper are: 1) it provides a bottom-up model for analyzing changes of a nation's product space structure; and 2) it also provides both qualitative and quantitative analysis methods for a nation's product space structure.