Introducing a process framework for implementing models of large-scale real-world systems in software

Abstract

The analysis and modeling of difficult, complex systems of the real world is nowadays feasible because of major advances in the computer science field and in mathematical models and theories that provide a reliable study framework of nonlinear, nondeterministic behavior in nature and society. The development of software products for the study of this behavior is growing fast, aiming at completing the required scientific framework and at providing a computationally powerful environment. A software process for modeling such large-scale, complex systems is proposed and analyzed in the present work, and a consistent methodology is described through a series of steps for handling the peculiarities and difficulties observed in these systems. The proposed framework is demonstrated through modeling and software implementation of a system that serves as a predictor of the Greek currency market. The methodology is quite simple and straightforward, while its use can be easily extended to a wide variety of complex systems analysis and modeling. Copyright © 2004 John Wiley & Sons, Ltd.