ABSTRACT
It is widely accepted in the geographical community that agent-based simulation is a useful technique to model dynamics and change encapsulated in real-world processes (Batty and Jiang, 1999). Agent-based systems have been found useful for modelling geographical phenomenon primarily because they provide the possibility to model group interactions and behaviours that emerge from individual properties (Dibble, 1996). This emergent property in agent-based models (ABM), since first fully exploited and demonstrated in the model sugarscape (Epstein and Axtell, 1996) has been seen as a revolutionary development for social science
— #2
(Bankes, 2002). However, behavioural dynamics include not only individual and group behaviours but also interplay of space and time to determine the spatial behaviour and decision-making in a geographical environment. Behavioural dynamics is complex because of the inherent complexity in the individual cognitive capabilities combined with the qualitative nature of space-time conceptualisations that result from these individual encodings. The capabilities of agent-based methods for modelling behavioural dynamics have not been carefully investigated yet. Although previous work has examined the role that these models play in the theoretical emergence of social constructs (O’Sullivan and Haklay, 2000), the focus of interest within this chapter is on the assessment and reconstruction of existing modelling tools within an accepted theoretical framework. To ensure that the theoretical structure emerging from a model is understood and reviewed within the framework of constraints and assumptions that an agent-based technique offers, it is crucial to be aware of the methodological and analytical foundations and underlying assumptions of an ABM.
