Page 37 -

30 P. Davidsson and H. Verhagen

behaviour and decision-making). Several models have been based on theories
from economy, social psychology, sociology, etc. Guye-Vuillème (2004)pro-
vides an example of this with his agent-based model for simulating human
interaction in a virtual reality environment. The model is based on sociological
concepts such as roles, values, and norms and motivational theories from social
psychology to simulate persons with social identities and relationships.

In most simulation studies, the behaviour of the individuals is static in the sense
that decision rules or reasoning mechanisms do not change during the simulation.
However, human beings and most animals do have an ability to adapt and learn. To
model dynamic behaviour of individuals through learning/adaptation can be done in
many ways. For instance, both ACT-R and Soar have learning built in. Other types
of learning include the internal modelling of individuals (or the environment) where
the models are updated more or less continuously.
Finally, there are some more general aspects to consider when modelling
individuals. One such aspect is whether all agents share the same behaviour or
whether they behave differently, in other words, representation of behaviour is
either individual or uniform. Another general aspect is the number of individuals
modelled, i.e. the size of the model, which may vary from a few individuals to
billions of individuals. Moreover, the population of individuals could be either
static or dynamic. In dynamic populations, changes in the population are modelled,
typically births and deaths.

3.4.2 Interaction Between Individuals

In dynamic micro-simulation, simulated individuals are considered in isolation
without regard to their interaction with others. However, in many situations, the
interaction between individuals is crucial for the behaviour at system level. In
such cases, better results will be achieved if the interaction between individuals
was included in the model. Two important aspects of interaction are (a) who
is interacting with whom, i.e. the interaction topology, and (b) the form of this
interaction.
A basic form of interaction is physical interaction or interaction based on
spatial proximity. As we have seen, this is used in simulations based on cellular
automata, e.g. in the well-known Game of Life (Gardner 1970). The state of an
individual is determined by how many of its neighbours are alive. Inspired by
this, work researchers developed more refined models, often modelling the social
behaviour of groups of animals or artificial creatures. One example is the BOID
model by Reynolds (1987), which simulates coordinated animal motion such as bird
flocks and fish schools in order to study emergent phenomena. In these examples,
the interaction topology is limited to the individuals immediately surrounding an
individual. In other cases, as we will see below, the interaction topology is defined
more generally in terms of a (social) network. Such a network can be either static,