<s>
CHREST	B-Architecture
(	O
Chunk	B-Architecture
Hierarchy	I-Architecture
and	I-Architecture
REtrieval	I-Architecture
STructures	I-Architecture
)	O
is	O
a	O
symbolic	O
cognitive	B-Architecture
architecture	I-Architecture
based	O
on	O
the	O
concepts	O
of	O
limited	O
attention	O
,	O
limited	O
short-term	O
memories	O
,	O
and	O
chunking	O
.	O
</s>
<s>
This	O
can	O
be	O
contrasted	O
with	O
Soar	B-Architecture
and	O
ACT-R	B-Language
,	O
two	O
other	O
cognitive	B-Architecture
architectures	I-Architecture
,	O
which	O
use	O
productions	O
for	O
representing	O
knowledge	O
.	O
</s>
<s>
CHREST	B-Architecture
has	O
often	O
been	O
used	O
to	O
model	O
learning	O
using	O
large	O
corpora	O
of	O
stimuli	O
representative	O
of	O
the	O
domain	O
,	O
such	O
as	O
chess	B-Application
games	O
for	O
the	O
simulation	O
of	O
chess	B-Application
expertise	O
or	O
child-directed	O
speech	O
for	O
the	O
simulation	O
of	O
children	O
's	O
development	O
of	O
language	O
.	O
</s>
<s>
In	O
this	O
respect	O
,	O
the	O
simulations	O
carried	O
out	O
with	O
CHREST	B-Architecture
have	O
a	O
flavour	O
closer	O
to	O
those	O
carried	O
out	O
with	O
connectionist	O
models	O
than	O
with	O
traditional	O
symbolic	O
models	O
.	O
</s>
<s>
CHREST	B-Architecture
stores	O
its	O
memories	O
in	O
a	O
chunking	O
network	O
,	O
a	O
tree-like	O
structure	O
that	O
connects	O
and	O
stores	O
knowledge	O
and	O
information	O
acquired	O
,	O
allowing	O
for	O
greater	O
efficiency	O
in	O
information	O
processing	O
.	O
</s>
<s>
CHREST	B-Architecture
is	O
developed	O
by	O
Fernand	O
Gobet	O
at	O
Brunel	O
University	O
and	O
Peter	O
C	O
.	O
Lane	O
at	O
the	O
University	O
of	O
Hertfordshire	O
.	O
</s>
<s>
It	O
is	O
the	O
successor	O
of	O
EPAM	B-Architecture
,	O
a	O
cognitive	O
model	O
originally	O
developed	O
by	O
Herbert	O
A	O
.	O
Simon	O
and	O
Edward	O
Feigenbaum	O
.	O
</s>
<s>
Chunks	O
in	O
CHREST	B-Architecture
are	O
referenced	O
in	O
short-term	O
memory	O
while	O
being	O
held	O
in	O
long	O
term	O
memory	O
,	O
often	O
recognised	O
through	O
neural	O
categorial	O
perception	O
involving	O
discrimination	O
.	O
</s>
<s>
In	O
much	O
similarity	O
to	O
EPAM	B-Architecture
,	O
chunks	O
in	O
cognition	O
learning	O
in	O
long-term	O
memory	O
are	O
acquired	O
as	O
a	O
“	O
network	O
of	O
nodes	O
”	O
,	O
and	O
are	O
interconnected	O
by	O
the	O
similarity	O
of	O
their	O
contents	O
and	O
are	O
depicted	O
as	O
a	O
discrimination	O
network	O
,	O
storing	O
and	O
sorting	O
chunks	O
in	O
the	O
network	O
.	O
</s>
<s>
Chunks	O
are	O
essentially	O
“	O
clusters	O
of	O
information	O
that	O
can	O
be	O
used	O
as	O
units	O
of	O
perception	O
”	O
,	O
thus	O
when	O
applied	O
in	O
situations	O
of	O
chess	B-Application
play	O
,	O
fragments	O
and	O
sections	O
of	O
chess	B-Application
positions	O
will	O
be	O
used	O
as	O
the	O
stimuli	O
fed	O
to	O
the	O
system	O
.	O
</s>
<s>
For	O
example	O
,	O
during	O
the	O
learning	O
phase	O
of	O
the	O
chess	B-Application
simulations	O
,	O
the	O
program	O
incrementally	O
acquires	O
chunks	O
and	O
templates	O
by	O
scanning	O
a	O
large	O
database	O
of	O
positions	O
taken	O
from	O
master-level	O
games	O
.	O
</s>
<s>
Taken	O
together	O
with	O
the	O
presence	O
of	O
time	O
and	O
capacity	O
parameters	O
,	O
this	O
enables	O
CHREST	B-Architecture
to	O
make	O
unambiguous	O
and	O
quantitative	O
predictions	O
.	O
</s>
<s>
CHREST	B-Architecture
’s	O
notability	O
lies	O
in	O
the	O
significance	O
placed	O
on	O
the	O
perception	O
process	O
.	O
</s>
<s>
This	O
phenomenon	O
is	O
closely	O
observed	O
in	O
chess	B-Application
experiments	O
,	O
where	O
perception	O
and	O
eye	O
movements	O
are	O
closely	O
associated	O
,	O
while	O
also	O
being	O
proportionate	O
to	O
attention	O
span	O
.	O
</s>
<s>
In	O
the	O
case	O
of	O
chess	B-Application
experiments	O
,	O
perception	O
is	O
equated	O
with	O
eye	O
movements	O
(	O
which	O
are	O
approximately	O
correspondent	O
to	O
attention	O
)	O
,	O
which	O
are	O
directed	O
by	O
chunks	O
held	O
in	O
memory	O
and	O
heuristics	O
.	O
</s>
<s>
Models	O
based	O
on	O
CHREST	B-Architecture
have	O
been	O
used	O
,	O
among	O
other	O
things	O
,	O
to	O
simulate	O
data	O
on	O
the	O
acquisition	O
of	O
chess	B-Application
expertise	O
from	O
novice	O
to	O
grandmaster	O
,	O
children	O
's	O
acquisition	O
of	O
vocabulary	O
,	O
children	O
's	O
acquisition	O
of	O
syntactic	O
structures	O
,	O
and	O
concept	O
formation	O
.	O
</s>
<s>
A	O
glaring	O
limitation	O
of	O
the	O
CHREST	B-Architecture
theory	O
is	O
as	O
proposed	O
by	O
Herbert	O
Simon	O
.	O
</s>
<s>
Simon	O
concluded	O
models	O
that	O
attempted	O
to	O
simulate	O
functioning	O
cognition	O
in	O
humans	O
must	O
not	O
assume	O
properties	O
that	O
may	O
be	O
unrealistic	O
for	O
a	O
human	O
,	O
thus	O
the	O
CHREST	B-Architecture
model	O
is	O
limited	O
by	O
the	O
parameters	O
of	O
human	O
abilities	O
understood	O
to	O
the	O
current	O
extent	O
of	O
cognitive	O
psychology	O
.	O
</s>
<s>
Time-related	O
variables	O
are	O
commonly	O
used	O
in	O
CHREST	B-Architecture
and	O
its	O
subsequent	O
simulations	O
,	O
such	O
as	O
the	O
main	O
limiting	O
factor	O
of	O
visual	O
short-term	O
memory	O
being	O
restricted	O
.	O
</s>
<s>
The	O
chess	B-Application
domain	O
has	O
long	O
been	O
a	O
standardised	O
testing	O
protocol	O
for	O
studies	O
involving	O
perception	O
,	O
psychology	O
,	O
cognition	O
,	O
and	O
human	O
and	O
artificial	B-Application
intelligence	I-Application
.	I-Application
</s>
<s>
The	O
comprehensive	O
use	O
of	O
chess	B-Application
play	O
and	O
chess	B-Application
mechanisms	O
has	O
been	O
compared	O
to	O
the	O
metaphor	O
of	O
the	O
use	O
of	O
‘	O
drosophila’	O
,	O
the	O
“	O
organism	O
of	O
choice	O
”	O
for	O
research	O
in	O
biological	O
and	O
chemical	O
industries	O
.	O
</s>
<s>
Similarities	O
between	O
the	O
domination	O
of	O
chess	B-Application
used	O
as	O
an	O
experimental	O
hotbed	O
in	O
the	O
field	O
of	O
cognitive	O
and	O
computer	O
sciences	O
and	O
the	O
use	O
of	O
drosophila	O
in	O
genetic	O
sciences	O
research	O
have	O
been	O
drawn	O
up	O
as	O
chess	B-Application
has	O
notably	O
been	O
identified	O
as	O
a	O
“	O
representative	O
measure	O
”	O
of	O
cognition	O
and	O
intelligence	O
in	O
both	O
humans	O
and	O
computers	O
.	O
</s>
<s>
Common	O
applications	O
and	O
simulations	O
of	O
the	O
CHREST	B-Architecture
theory	O
have	O
been	O
carried	O
out	O
extensively	O
in	O
the	O
past	O
within	O
the	O
context	O
of	O
chess	B-Application
play	O
.	O
</s>
<s>
Varying	O
networks	O
of	O
nodes	O
(	O
or	O
chunks	O
)	O
of	O
different	O
sizes	O
are	O
then	O
created	O
,	O
which	O
allows	O
for	O
simulations	O
of	O
chess	B-Application
play	O
across	O
diverse	O
levels	O
of	O
skill	O
.	O
</s>
<s>
Parameters	O
of	O
time	O
and	O
human	O
capacity	O
are	O
taken	O
into	O
account	O
,	O
thus	O
ideally	O
creating	O
circumstances	O
where	O
CHREST	B-Architecture
is	O
able	O
to	O
quantitatively	O
predict	O
unambiguous	O
outcomes	O
(	O
Gobet	O
and	O
Lane	O
;	O
Gobet	O
)	O
.	O
</s>
<s>
Additional	O
research	O
credited	O
to	O
Adriaan	O
de	O
Groot	O
and	O
Herbert	O
Simon	O
specifically	O
in	O
the	O
domain	O
of	O
chess	B-Application
accounted	O
for	O
significant	O
quantities	O
of	O
psychological	O
data	O
,	O
with	O
a	O
strong	O
focus	O
on	O
the	O
memory	O
of	O
chess	B-Application
players	O
.	O
</s>
<s>
Prior	O
to	O
de	O
Groot	O
and	O
Simon	O
’s	O
theories	O
and	O
implementation	O
,	O
the	O
standard	O
paradigm	O
for	O
experimentation	O
in	O
chess	B-Application
play	O
and	O
chess	B-Application
research	O
typically	O
consists	O
of	O
illustrating	O
a	O
chess	B-Application
position	O
to	O
a	O
subject	O
for	O
a	O
short	O
period	O
of	O
time	O
,	O
usually	O
for	O
5	O
seconds	O
,	O
then	O
asking	O
subjects	O
to	O
recreate	O
the	O
position	O
.	O
</s>
<s>
In	O
the	O
domain	O
of	O
perception	O
,	O
simulations	O
of	O
eye	O
movement	O
during	O
the	O
initial	O
5	O
seconds	O
of	O
illustrating	O
a	O
chess	B-Application
position	O
,	O
as	O
well	O
as	O
recognition	O
of	O
templates	O
and	O
chunks	O
have	O
been	O
completed	O
using	O
CHREST	B-Architecture
.	O
</s>
<s>
CHREST	B-Architecture
also	O
accounts	O
for	O
the	O
outcome	O
when	O
presented	O
with	O
varying	O
modifications	O
and	O
randomisation	O
of	O
positions	O
,	O
the	O
significance	O
of	O
time	O
spent	O
illustrating	O
and	O
presenting	O
each	O
position	O
,	O
and	O
the	O
categorisation	O
of	O
the	O
errors	O
made	O
and	O
chunks	O
replaced	O
in	O
the	O
network	O
across	O
varying	O
skill	O
levels	O
from	O
novice-level	O
players	O
to	O
grandmasters	O
.	O
</s>
<s>
Extensive	O
research	O
has	O
been	O
conducted	O
by	O
N	O
Charness	O
on	O
chess	B-Application
and	O
general	O
expertise	O
,	O
problem-solving	O
strategies	O
and	O
memorisation	O
by	O
population	O
groups	O
of	O
different	O
ages	O
.	O
</s>
<s>
Tests	O
for	O
memorisation	O
and	O
recall	O
revealed	O
that	O
younger	O
players	O
performed	O
better	O
relative	O
to	O
older	O
players	O
when	O
presented	O
with	O
varying	O
chess	B-Application
positions	O
.	O
</s>
<s>
Prior	O
to	O
de	O
Groot	O
and	O
Simon	O
’s	O
theories	O
and	O
implementation	O
,	O
the	O
standard	O
paradigm	O
for	O
experimentation	O
in	O
chess	B-Application
play	O
and	O
chess	B-Application
research	O
typically	O
consists	O
of	O
illustrating	O
a	O
chess	B-Application
position	O
to	O
a	O
subject	O
for	O
a	O
short	O
period	O
of	O
time	O
,	O
usually	O
for	O
5	O
seconds	O
,	O
then	O
asking	O
subjects	O
to	O
recreate	O
the	O
position	O
.	O
</s>
<s>
Though	O
this	O
methodology	O
has	O
generated	O
a	O
substantial	O
amount	O
of	O
high-level	O
models	O
addressing	O
memory	O
and	O
cognition	O
in	O
chess	B-Application
play	O
,	O
exampled	O
by	O
the	O
works	O
of	O
Dennis	O
Holding	O
,	O
there	O
remains	O
a	O
scarcity	O
of	O
models	O
that	O
further	O
detail	O
memory	O
use	O
in	O
chess	B-Application
,	O
with	O
the	O
exemption	O
of	O
MAPP	O
developed	O
by	O
Chase	O
and	O
Simon	O
,	O
later	O
implemented	O
by	O
Simon	O
and	O
Gilmartin	O
.	O
</s>
