Neural Networks:

Neural networks represent a concept for a computer architecture
that will more accurately emulate human thought processes. 
Neural networks in the brain share the information base and the
processing demands, working in parallel (simultaneously) to share
and reinforce patterns of information structure, including
associative links, experiential clues, and response memories. 
Although working at a slower rate than the electronic serial
computers, each neuron is, in itself, an information processor
(Clarkson, 1989).  If computers can accommodate an effective
network of associative links (shared associations), with many
processors dividing the computational tasks among themselves
while running in parallel, then the computer will more closely
mimic human thought processes.  Additionally, such systems could
be self-informing, developing new associative links and creating
new patterns of relationships in the knowledge base, and
reinforcing old ones through repeated recognition and
utilization.  

This technology has implications for the capability of the
computer as a true decision making machine (in addition to
serving as an information source).  Computers based on neural
networks will be capable of discretionary tasks, responding to
particular environmental conditions and context.  Powerful
monitoring systems with the capability to intervene in the
resource protection process may be possible.  Analytic systems
for remotely sensed data concerning buildings and material
components could become part of a facility's environmental
control system. The predictive modeling capabilities of neural
network systems would be able to indicate the effects of weather,
pollutants, visitation demands, planned maintenance activities,
and design decisions on the integrity of a cultural resource.

<ffile32>   Parallel Processing
<cfile25>   Hypertext Introduction