Information Technology Reference
In-Depth Information
DW and HK models but allow for a better understanding of the approximations
that need to be made in order to better understand the
d
and
parameters of
the DW and HK models.
The drawback of the above discussed models is that, while implemented at
agent level they are limited in their explanation of the effect of space on the diffu-
sion of opinion. Factors like location, shape and size of land use affect the opinions
of actors. For example certain areas might be considered “non-negotiable” by ac-
tors because the have a strong emotional or economic meaning to an actor. This
means that, to be able to explore opinion dynamics in a spatial environment,
current opinion dynamics models need to be adapted to represent the effects of
space on the rate of change and distribution of opinions both in space as well as
between actors.
In this paper an agent based opinion dynamics model, based on the DW
approach is proposed that explicitly represents the additional complexity of a
spatial decision making systems. For most spatial problems the opinion space
can be considered multi-dimensional as the opinion dynamics not only depend
on time but also its location within a 2 dimensional geographic space. Moreover
opinions are not only influenced by previous time steps but in addition also by
the opinions agents have about other locations.
3 A Model for Spatial Opinion Dynamics
The proposed ABM represents a category of spatial decision-making problems
where multiple actors having multiple desires and goals regarding the future
development of an area have to generate a joint vision for future development
i.e. a spatial plan. Each agent tries to realize its individual goals as good as
possible within the context of goals of the other agents. The proposed model
currently takes only one issue into account. This can be for example an allocation
problem dealing with the question: where to realize a land use function (e.g. new
urban areas). The model focuses on the dynamics of the opinions assuming that
there is an intitial opinion for each agent, resulting from decision-making at
individual level. As such it builds on previous work which described an agent
based approach where land use change is purely driven by rational processes
such as voting [20, 21]. In the proposed model, the dynamics of opinions are
characterized by the following aspects:
-
the initial opinion of an agent about a location;
-
the social influence and;
-
the spatial influence.
Including these three aspects the dynamics of opinion
O
k,
(
t
+1)
of agent
k
at time
t
+ 1 in a social-spatial system can be defined as:
O
k,
(
t
+1)
=
f
(
Oi
(
k,g,t
)
, Q
(
k,g,t
)
, Q
(
k,g,t
)
)
(4)
Where
Oi
k,g
is the the individual opinion of agent
k
for land use
g
,
Qs
k,g
the
effect of the social influence of agent
k
for land use
g
,and
Qg
k,g
is effect of
