Information Technology Reference
In-Depth Information
11.4 Computational Modeling of Agents and Artifacts
11.4.1 Products
Products are constructed as a graph structure with six nodes. During the construction
process, edges between the nodes are formed at random, creating a single connected
graph with between 5 and 14 edges. There are 112 unique six node connected graphs,
yielding a design space for agent exploration that is small enough to be tractable for
enumeration and complete analysis.
Products have both surface characteristics and functional characteristics. During
their search and evaluation process, Consumers can only sense and perceive a
Product's surface characteristics (its 'signature'). The functional characteristics are
only experienced through the process of consumption. During consumption, Con-
sumers gain utility based on the functional characteristics, relative to expectations.
Higher than expected evaluations of functional characteristics yield positive utility,
while lower than expected evaluations of functional characteristics yield negative
utility. The surface characteristics of Products are related to their functional charac-
teristics, but not identical. Consumers cannot directly perceive the utility of Products.
The Product's utility is a function of its topology, while its external appearance
is a function of its physical layout. Physical layout (Fig.
11.1
) is constructed on a
circle with a relaxation method to equalize the length of edges. Distance from the
centroid of this layout produces a signature in the form of a six-element vector. Four
more elements are added based on other surface characteristics of the graph, for a
total of ten elements in a signature vector.
Three dimensions of Consumer utility have been defined (Fig.
11.2
) in a way that
might be analogous to performance dimensions in a manufactured component. The
first is the number of edges, which ranges between 5 and 15, and might be
analogous to a property like weight or density. The second is mean clustering
coefficient, which might be analogous to a property like stiffness/flexibility. The
third dimension is the maximum distance between nodes in a spring-electrical
embedding. This might be analogous to maximum span distance.
The construction process for Products yields a non-obvious relationship between
the surface characteristics and the functional characteristics of Products. Products
that are close in surface characteristics (i.e. close in Value Space) might have
different utilities. This allows for a somewhat rugged landscapes.
a
b
c
d
e
signature:
other characteristics:
s
7
= max degrees
s
8
= variance in degrees
s
9
= # cycles
s
10
= graph diameter
distances from centroid =
{s
1
, s
2
, s
3
, s
4
, s
5
, s
6
}
Fig. 11.1 Generation of surface characteristics for Products. (a) random placement of nodes on a
unit circle and random generation of edges; (b) identification of a centroid; (c) and (d) distances of
nodes from centroid determines the signature; (e) other physical characteristics available to
Consumers through sensation and perception
Search WWH ::
Custom Search