Information Technology Reference
In-Depth Information
9.7 SOAR Based on Memory Chunks
By the end of 1950s, a model of storage structure was invented by means of
using one kind of signals to mark the other signals in neuron simulation. This is
the earlier concept of chunks. The chess master kept memory chunks about
experiences of playing chess under different circumstances in mind. In the early
of 1980s, Newell and Rosenbloom proposed that the system performance can be
improved by acquiring knowledge of model problem in task environment and
memory chunks can be regarded as the simulation foundation of human action.
By means of observing problem solving and acquiring experience memory chunk,
the complex process of each sub-goal is substituted and thus ameliorates the
speed of the problem solving of the system, thereafter laid a solid foundation for
empirical learning.
In 1986, J.E. Laird from the University of Michigan, Paul S. Rosenbloom
from the University of Stanford and A. Newell from Carnegie Mellon University
developed SOAR system(Laird et al., 1986), whose learning mechanism is to
learn general control knowledge under the guidance of outside expert. The outer
guidance can be direct, or an intuitionistic simple question. The system converts
the high level information from outer expert into inner presentations and learns to
search the memory chunk(Golding et al., 1987). Figure 9.10 presents the
architecture of SOAR.
The processing configuration is composed of production memory and
decision process. The production memory contains production rule, which can be
used for searching control decision. The first step is detailed refinement; all the
rules are referred to working memory in order to decide the priorities and which
context should be changed and how to change. The second step is to decide the
segment and goal that needs to be revised in the context stack.
Search WWH ::




Custom Search