Geoexpression: A Petri network framework for representing geographic process concurrency

Abstract

This article examines important characteristics missing from state-of-the-art representations of geographic processes with regard to concurrency. We investigate: what geographic process concurrency is and what opportunities for improved computational models exist by explicitly representing geographic process concurrency in a way that is mutually understandable by the human and the machine. Unapt representations of geographic process concurrency can lead to non-deterministic geographic dynamic modeling outcomes, excessive cognitive burdens when reasoning about how concurrent processes interact, and even inconsistent results. We use a geographic dynamic modeling example to reveal the existing research gap, and to demonstrate that graphical languages can be used AbstractWe present a novel framework called geoexpression for representing geographic process concurrency and discuss its implications for geographic dynamic modeling. Unapt representations of geographic process concurrency can lead to non-deterministic geographic dynamic modeling outcomes, excessive cognitive burdens when reasoning about how concurrent processes interact, and even inconsistent results. After demonstrating the importance of geographic process concurrency, we examine how the characteristics of geographic process concurrency are missing from other state-of-the-art geographic process representations. The geoexpression framework adopts Petri networks to allow researchers to understand how to relate geographic process concurrency to observed modeling patterns. Of interest are the real-world geographic patterns formed by concurrent geographic processes.

A Concurrent Entity Component System for Geographical Wildlife Epidemiological Modeling

“…Once the processes have been enumerated, Petri networks are used to represent their order and interactions graphically. As a graphical language, Petri networks are well suited to identify regions of latent concurrency that can be represented and exploited in computation (Davis and Wang 2018).…”

“…These investigations evolved to discern how cyberinfrastructure can serve as a basis for further inquiry toward the influence of space-time pattern representation using epidemiological modeling as a case study (Shook and Wang 2015). The next logical step, explored first in the work of Davis and Wang (2018) and now here, is the investigation of how cyberinfrastructure motivates the need to expand the theory of geographic process representation; specifically, to understand and exploit geographic process concurrency.…”

“…Geographical process concurrency is the decomposability of geographical processes into a structure of partially orderable components that take place within spatial and temporal configurations to represent geospatial dynamics. Once this system of concurrent processes is structured graphically, as examined by Davis and Wang's (2018) geoexpression, it can be used to exploit computationally performant software to represent latent concurrency in geographical processes.…”

“…Research has focused heavily on the representation of the pattern and very little on the representation of the process (Goodchild, Yuan, and Cova 2007). However, as introduced in Davis and Wang (2018), cyberinfrastructure has motivated the need to explore a new process representation theory that captures and exploits the notion of geographical process concurrency. Geographical process concurrency is an underexplored frontier in geographical theories; especially as it relates to the representation of processes using computational machines.…”

“…The geoexpression is a framework that structures geographical process concurrency using a graphical representation to express the relationships between a set of related geographical processes in a way that is mutually understandable by a human and computational machine. Davis and Wang (2018) conceptualize geographical process concurrency through geoexpression using a graphical language, Petri-net, to represent the concurrent execution order of a set of processes. Once the concurrent execution order is structured, it is feasible to identify regions within the model and software that are conducive to parallel execution, which cyberinfrastructure is well suited to support.…”