A context-driven representation for spatiotemporally-explicit agent-based models
The representation of contextual information poses a unique challenge in the development of agent-based models of complex adaptive spatial systems. This representation is particularly important for the modeling of agent-agent and agent-environment interactions. In the research presented here, we developed and implemented a context-driven model that supports the explicit representation of situation-dependent information for agent-based decision-making within spatiotemporally explicit environments. This context-driven model is part of a generic geocomputational framework designed to model geographically-aware intelligent agents. The context model is based on a typology of context elements, including identity, activity, space, and time. Contextual information is organized, modeled, and used by agents to enhance their problem-solving capabilities. In particular, information associated with the adaptive decision-making of agents is explicitly taken into account in this framework to facilitate the modeling of agent-based learning within geographic systems. We examine the utility of context-driven modeling via an agent-based model of elk movement. Elk in this model are represented as geographically-aware intelligent agents that learn to adapt to landscape dynamics. Context is represented as the internal and external stimuli of elk during migration. Experimental results suggest that the context-driven framework is well suited to modeling cognizant agents within complex adaptive spatial systems.