International Journal of Intelligence Science, 2011, 1, 25-34
doi:10.4236/ijis.2011.12004 Published Online October 2011 (
Copyright © 2011 SciRes. IJIS
Cognitive Cycle in Mind Model CAM*
Zhongzhi Shi, Xiaofeng Wang, Jinpeng Yue
The Key Laboratory of Intelligent Information Processing, Institute of Computing Technology,
Chinese Academy of Sciences, Beijing, China
Received October 2, 201 1; revised October 10, 2011; accepted October 15, 2011
Cognitive cycle is a basic procedure of mental activities in cognitive level. Human cognition consists of cas-
cading cycles of recurring brain events. This paper presents a cognitive cycle for the mind model CAM
(Consciousness and Memory). Each cognitive cycle perceives the current situation, through motivation phase
with reference to ongoing goals, and then composes internal or external action streams to reach the goals in
response. We use dynamic description logic which is an extended description logic with action to formalize
descriptions and algorithms of cognitive cycle. Two important algorithms, including hierarchical goal and
action composition, are proposed in the paper.
Keywords: Cognitive Cycle, Motivation, Action Composition, CAM
1. Introduction
Cognitive cycle is a basic procedure of mental activities
in cognitive level. Human cognition consists of cascade-
ing cycles of recurring brain events. In problem solving,
particular for production systems, solving cycles were
proposed. In the early 1980’s, SOAR was developed to
be a system that could support multiple problem solving
methods for many different problems [1]. In the mid
1980’s, Newell and many of his students began working
on SOAR as a candidate of unified theories of cognition
[2]. SOAR is a classic example of expert rule-based cog-
nitive architecture designed to model general intell igence
with a learning architecture that has been applied to do-
mains ranging from rapid, immediate tasks such as typ-
ing and video game interaction to long stretches of prob-
lem solving behavior. SOAR has also served as the
foundation for a detailed theory of sentence processing,
which models both the rapid on-line effects of semantics
and context, as well as subtle effects of syntactic struc-
ture on processing difficulty across several typologically
distinct languages.
The adaptive control of thought-rational (ACT-R)
model, developed mainly by Anderson [3], which is a
symbolic cognitive architecture aiming to explain how
the components of the mind work together to produce
coherent cognition. Coordination of the ACT-R modules
is achieved by a central production system shown in
Figure 1 [4].
ACT-R is a hybrid cognitive architecture. Its symbolic
structure is a production system; the subsymbolic struc-
ture is represented by a set of massively parallel proc-
esses that can be summarized by a number of mathe-
matical equations. The subsymbolic equations control
many of the symbolic processes. If several productions
match the state of the buffers, a subsymbolic utility
equation estimates the relative cost and benefit associ-
*This paper is supported by Key projects of National Natural Science
Foundation of China (No. 61035003, 60933004), National Natural
Science Foundation of China (No. 61072085, 60970088, 60903141),
ational Basic Research Pro
. Figure 1. ACT-R archit ecture [4].
ated with each production and decides to select for exe-
cution the production with the highest uti lity.
In practice ACT-R seems to be used more as a pro-
gramming framework for cognitive modeling than as an
AI system. One can easily use ACT-R to program mod-
els of specific human mental behaviors, which may then
be matched against psychological and neurobiological
data [5]. Recently Anderson and his colleagues investi-
gateACT-R under fMRI guide to execute algebra opera-
tions [6,7].
An agent perceives the environment through the sen-
sor and effect the environment with the actor to commu-
nicat e with e nvir onment [8]. A huma n bein g, if we see it
an agent, perceives environment with eyes, ear, nose, etc.
It affects the environment using hands, legs etc. A robot
agent u s ual l y uses some c a me r as fo r get ti ng en vir o n me nt
information and the motor is their actor. A software
agent uses codes as their sensor and actor. In 1993, Sho-
ham proposed agent-oriented programming in terms of
BDI agent architectures [9]. The Belief-Desire-Inten-
tion (BDI) agent model is an event-driven execution
model providing both reactive and proactive behavior.
The BDI agent model is built on a simplified view of
human intelli gence. In it, agen t s have a view of the world
(Beliefs), certain goals they wish to achieve (Desires),
and they form Plans (Intentions) to act on these using
their accumulated experience. Agents based on the BDI
model are at a level of abstraction closer to normal hu-
man experience [10].
In AI terms, Beliefs represent knowledge of the world.
However, in computational terms, Beliefs are just some
way of representing the state of the world, be it as the
value of a variable, a relational database, or symbolic
expressions in predicate calculus. Beliefs are essential
because the world is dynamic, and the system only has a
local view of the world. Moreover, as the system is re-
source bounded, it is desirable to cache important infor-
mation rather than recompute it from base perceptual
data. Desires form another essential component of sys-
tem state. Again, in computational terms, a Goal may
simply be the value of a variable, a record structure, or a
symbolic expression in some logic. The important point
is that a Goal represents some desired end states. The
underlying semantics for Goals, irrespective of how they
are represented computationally, should reflect some
logic of desire. In the AI literature, Intentions represent
the third necessary component of system state. Computa-
tionally, Intentions may simply be a set of executing
threads in a process that can be appropriately interrupted
upon receiving feedback from the possibly changing
world. The basic components of a system designed for a
dynamic, uncertain world should include some represen-
tation of Beliefs, Desires, Intentions and Plans, or what
has come to be called a BDI agent [11].
CLARION is a hybrid architecture that combines a
symbolic component for reasoning on explicit know-
ledge’ with a connectionist component for managing
implicit knowledge. [12] Learning of implicit knowledge
may be done via neural net, reinforcement learning, or
other methods. The integration of symbolic and sub-
symbolic methods is powerful, but a great deal is still
missing such as episodic knowledge and learning and
creativity. Learning in the symbolic and subsymbolic
portions is carried out separately rather than dynamically
coupled, minimizing “cognitive synergy” effects. CLA-
RION consists of a number of distinct subsystems, each
of which contains a dual representational structure, in-
cluding a “rules and chunks” symbolic knowledge store
somewhat similar to ACT-R, and a neural net kno wledge
store embodying implicit knowledge. The CLARION
architecture is shown in Figure 2. It contains main sub-
systems as follows:
1) An action-centered subsystem to control actions;
2) A non-action-centered subsystem to maintain gen-
eral knowledge;
3) A motivational subsystem to provide underlying
motivations for perception, action, and cognition;
4) A meta-cognitive subsystem to monitor, direct, and
modify the operations of all the other subsystems.
Motivational dynamics is an essential part of human or
animal behaviors. In CLARION Sun proposed a motiva-
tional subsyste m sho wn in Figure 3 [13]. In this subsys-
tem, the goal structure constitutes an explicit representa-
tion of motivations, and drives an implicit one. The
mapping between the state of the world, for instance,
stimuli as perceived by a cognitive agent, and the sensing
of various perceived deficits, and the strengths of various
drives can be implemented, in accordance with the afore-
specified value ranges and relatio ns, by back-propagation
networks. The networks identify relevant features from
raw sensory input. The output of such a network may be
the strengths of drives.
The LIDA architecture developed by Stan Franklin
and his colleagues is based on the concept of the cogni-
tive cycle—a notion that is important to the brain [14,
15]. Each cognitive cycle the LIDA agent first makes
sense of its current situation as best as it can. It then de-
cides what portion of this situation is most in need of
attention. Broadcasting this portion, the current contents
of consciousness, enables the agent to finally choose an
appropriate action and execute it.
Autonomous agents cope with their changing envi-
ronment by their continuous, cyclic chores of “perceive-
understand-act”. LIDA’s cognitive cycle is the cycle of
refined cognitive processes that bring about the appro-
priate action for specific situation. As Franklin and Baars
Copyright © 2011 SciRes. IJIS
Copyright © 2011 SciRes. IJIS
Figure 2. CLARION architecture [12].
Cognitive cycle should reflect agent interaction with
environment, which is sensory outside environment,
through deliberation, and then effect environment. Cog-
nitive cycle is a whole procedure from perception to be-
havior execution. In this paper we propose a cognitive
cycle in mind model CAM which consists of three
phases: perception, motivation and action composition.
Dynamic description logic is used for formalizing de-
scriptions and algorithms. Two important algorithms,
including hierarchical goal and action composition, are
proposed in the paper.
Next section will introduce mind model CAM. The
CAM cognitive cycle will be discussed in Section 3. Fi-
nally, conclusion and perspective will be given.
Figure 3. Structure of the motivational subsystem [13].
put it “A cognitive cycle can be thought of as a moment
of cognition—a cognitive moment; higher-level cogni-
tive processes are composed of many of these cognitive
cycles, each a cognitive atom.” This metaphor is to say
that the steps in a cognitive cycle correspond to the vari-
ous sub-atomic particles in an atom. Since the LIDA
architecture is composed of several specialized mecha-
nisms, a continual process that causes the functional in-
teraction among the various compo nents is essential. T he
cognitive cycle as such is an iterative, cyclical, continu-
ally active process that brings about the interplay among
the various components of the architecture [16].
2. Mind Model CAM
Mind could be defined as: “That which thinks, reasons,
perceives, wills, and feels. The mind now appears in no
way separate from the brain. In neuroscience, there is no
duality between the mind and body. They are one.” in
Medical Dictionary [17]. A mind model is intended to b e
an explanation of how some aspect of cognition is ac-
complished by a set of primitive computational processes.
A model performs a specific cognitive task or class of
tasks and produces behavior that constitutes a set of pre-
dictions that can be compared to data from human per-
formance. Task domains that have received considerable
attention include problem solving, language comprehen-
sion, memory tasks, and human-device interaction.
A new mind model called Consciousness and Memory
(CAM) is proposed by Intelligence Science Laboratory
of Institute of Computing Technology [18]. Figure 4
shows the architecture of CAM model which consists of
three main parts, consciousness, memory and high level
cogniti ve func tions. The co nscious ness po ssesse s a set of
planning schemes which arrange the components of
CAM to accomplish different cognitive tasks. The mem-
ory part contains three types of memory which are long
term memory, short term memory and working memory.
The high level cognitive function part includes event
detection, action execution etc.
2.1. Semantic Memory
In semantic memorythat is, conceptual memory in
Figure 4, we use ontology to specify a conceptualization
of a domain in terms of concepts, attributes, and relations.
The concepts provide model entities of interest in the
domain. They are typically organized into a taxonomy
tree where each node represents a concept and each con-
cept is a specialization of its parent. Each concept in a
taxonomy is associated with a set of instances. By the
taxonomy's definition, the instances of a concept are also
instances of an ancestor concept. Each concept is also
associated with a set of attributes. An ontology also de-
fines a set of relations among its concepts. Logic lan-
guage Dynamic description logic (DDL) proposed by
authors’ Lab defines ontology terminologies and expres-
siveness [19].
An example of semantic memory to describe animal
hierarchical category is given in Figure 5. We can de-
scribe it in DDL as follows:
Animal Bird Fish
Bird Canary Ostrich
Fish Shark Salmon
animal(x,y) ((HasSkin(x)
CanMoveAround (x)
Eats(x) Breathes(x)),
bird(x,y) ((HasWings(x) CanFly(x)
HasFeathers(x)), Bird(y))
An example of semantic memory to describe animal
hierarchical category is given in Figure 5. We can de-
scribe it in DDL as follows:
Animal Bird Fish
Bird Canary Ostrich
Fish Shark Salmon
animal(x,y) ((HasSkin(x)
CanMoveAround (x)
Eats(x) Breathes(x)),
bird(x,y) ((HasWings(x) CanFly(x)
HasFeathers(x)), Bird(y))
Figure 4. Architecture of CAM.
Copyright © 2011 SciRes. IJIS
Z. Z. SHI ET AL.29
Figure 5. Semantic memory [20].
2.2. Episodic Memory
Episodic memory is a category of long-term memory that
involves the recollection of specific events, situations
and experiences [21]. Nuxoll and Laird demonstrated
that an episodic memory can support an intelligent agent
to own a multitude of cognitive capabilities [22].
In CAM the episode is an elementary unit that stores
previous scene in episodic memory where an episode is
divided into two levels: one is an abstract level in terms
of logic, another is a primitive level shown in Figure 6.
Among them, episode is represented in the form of logic
symbol on the abstract level. The primitive level includes
perception information correlated to abstract level of the
described object.
In order to represent and organize perception of the
episode effectively, we adopt DDL to describe episode in
abstract level and ontology in primitive level. Object data
graph (ODG) is used to describe episode. Figure 7 de-
picts an ODG structure of film Waterloo bridge where
objects associate with other objects through URI in epi-
sode. Figure 7 shows us 3 objects: M2, W2, and film
Waterloo bridge. In addition, object W2 has worn a blue
skirt. The film also associates with two main roles M1,
W1 and among them W1 has worn a white coat.
In Soar, the retrieval of episode is modeled as a
case-based reasoning problem which finds solutions to
problems according previous experience [22]. We follow
this idea and build a case based system to retrieve the
episode according to the cues. To simplify the system,
we restrict the cue to be transitional sequence like epi-
sode. Then, the retrieval of episode is modeled as prob-
Figure 6. Example of episode.
lem of finding the episode that is most relevant to the cue.
As the abstract level episode can represent the content of
episode precisely, thus the matchmaking is only per-
formed between cue and abstract level episode. In CAM,
the transitional sequence is formally defined as possible
world sequence and whether the episode implies cue can
be inferred by the DDL based tableau algorithms [18].
3. CAM Cognitive Cycle
In mind model CAM we propose the cognitive cycle
shown in Figure 8. The CAM cognitive cycle depicts as
Perception-Motivatio n-Action Composition three phases.
Copyright © 2011 SciRes. IJIS
Copyright © 2011 SciRes. IJIS
Figure 7. Object data graph.
Figure 8. CAM cognitive cycle.
Perception phase is the process of attaining awareness of
the environment by sensory input. Using the incoming
percept and the residual contents of working memory, as
cues, local associations are automatically retrieved from
transient episodic memory and from declarative memory.
Motivation phase focuses on learners’ beliefs, expecta-
tions, and needs for order and understanding. According
to the impact factors of motivation, such as proportional
activation, opportunism, contiguity of action, persistence,
interruption, combination of preference we construct a
motivation subsystem. Action composition will compose
a group of actions through action selection, planning to
reach the end goal.
3.1. Perception
Perception phase is the process of attaining awareness or
understanding of the environment by organizing and
interpreting sensory information [23,24]. All perception
involves signals in the nervous system, which in turn
result from physical stimulation of the sense organs.
Sensory stimuli, external or internal, are received and
interpreted by perception producing the beginnings of
Awareness is the state or ability to perceive, to feel, or
to be conscious of events, objects or sensory patterns. In
this level of consciousness, sense data can be confirmed
by an observer without necessarily implying understand-
ing. More broadly, it is the state or quality of being
aware of something. In biological psychology, awareness
is defined as a human’s or an animal’s perception and
cognitive reaction to a condition or event.
3.2. Motivation
Motivation is a process that starts with a physiological or
psychological need that activates a behavior or a drive
that is aimed at a goal. There are several motivation
theories, such as Extrinsic motivation, Behaviorism,
Humanistic views of motivation, Cognitive theories of
Physiologist Maslow developed the hierarchy of needs
through his paper titled a Theory of Human Motivation
in 1943 [25]. Since it introduced to the public, the
Maslow’s hierarchy of needs theory has been made a
significant impact to the every life aspect in people’s life.
This theory can give people more spirit and motivation
so they can manage their life very well [26]. The
Maslow’s hierarchy of needs is describing the reality of
most people life experience accurately. The Maslow’s
Z. Z. SHI ET AL.31
hierarchy of needs theory is divided into five different
levels of basic needs, including physiological needs in
the lowest level, security needs in the second level, needs
of love, affection and ownership in the next le vel, esteem
needs in the fourt h level, and the last is se lf-actualization
needs in the top of hierarchy shown in Figure 9. Maslo w
actually was a humanistic psychologist who believed in
the human potential that human can struggle to reach the
success and look for the creativity in order to reach the
highest wisdom and also the logic think. From above we
can see that huma nistic views of moti vatio n foc us on th e
learner as a whole person and examine the relationships
among physical, emotional, intellectual, and aesthetic
Bach has proposed a framework for an extensible mo-
tivational system of cognitive agents, based on research
in psychology [27]. It draws on a finite set of pre-defined
drives, which relate to needs of the system. Goals are
established through reinforcement learning by interacting
with an environment. Bach also points out that all be-
havior of Psi agents is directed towards a goal situation,
which is characterized by a consumptive action satisfy-
ing one of the needs. Bach proposes hierarchy of agent
needs with three levels shown in Figure 10. The lowest
Figure 9. Hierarchy of ne eds.
Figure 10. Hierarchy of agent needs.
level is physiological needs, containing fuel and water,
intactness. Second level is cognitive needs, containing
certainty, competence, and aesthetics. The third level is
social needs, containing affiliation and supplication sig-
Bach uses Psi theory to define a possible solution for a
drive-based, poly-thematic motivational system. It can
reflect physiological needs, also addresses the establish-
ment of cognitive and social goals. Its straightforward
integration of needs allows adapting it quickly to differ-
ent environments and types of agents. They develop a
version of the model which has been successfully evalu-
ated against human performance in problem solving
games [28].
In the motivation phase of the mind model CAM, an
explicit goal may be set based on drives according to the
needs. A goal list consists of a number of goals which
can be described formally:
tt t
GGG Gat time t
Definition 1: A hierarchical goal which is a directed
acyclic graph (DAG) can be defined as a 3-tuples:
DAG = (P, E, )
P: a set of nodes
E: a set of edges which indicate the relation between
connected nodes
: partial order
Should be satisfied following conditions:
1) A particular node TopGoal is contained in P,
,, and p TopGoal
2) if 12
, then 12
3) if 12
 then 21
4) if
12 n
,p ,,pP
, and
,pEp ,pE
 ,
then 2n
,p E,,p,p E
 
Let a set of goals at time t be
, where, , let ,1
ttt t
iii i
GGG G, 1kn
, satisfy
, that is, no repeated goals in
. The algorithm for creating hierarchical
goals is given a s follows:
Algorithm 1: Hierarchical goals
Input: a set of goals and their
partial order
ttt t
iii i
Output: DAG which is a directed acyclic graph
1) Initialize DAG null;
Copyright © 2011 SciRes. IJIS
2) if is empty, then end and return DAG;
3) take
G from ,
k, update
4) if is empty, create root = new node (“root”),
generate i
e G
, and let
arentrootnode G;
5) call
insertroot G;
6) Goto 2.
insertroot G
1) let Children = children(root);
2) if Children is empty, create a new node
arentnode Groot, return;
3) take Child fr om Children, let
ChildrenChildren Child
4) if no partial order between Child. Concept and
then Goto 2;
5) if Child .Concept
G, then create a new node
nodeG, let
arent nrootode G, and
arent ChildnodeG,
arent Childroot
,, return;
6) call
insertChild Gi.
1) Func tio n children(node C) find all connected nodes
directly with node C;
2) Function parent(node C) find directed parent node
of node C;
3) Function new node(C) create a new node C;
4) Function parent(node(C)) = node(D) assign no de D
as a parent node to node C.
3.3. Action Composition
Action composition is the pro cess of constructi ng a com-
plex composite action from atomic actions to achieve a
specific task. Action composition can be divided into two
steps, first is action selection, i.e., select related action
from action library. Then selected actions are composed
together using a pla nning strategy.
The action selection chooses a single action from a
just instantiated action stream or possibly from a previ-
ously active stream. There are a lot of selection methods.
Most of them match goal and behavior based on similar-
ity criteria.
Planning offers a scalable and efficient approach for
action composition. It allows for a composition request
to be expressed in terms of goal conditions that specify a
set of constraints and preferences. We use dynamic de-
scription logic to describe definitions and algorithm for-
mally [19].
Definition 2 (Action) An action description is the
form of
1,, ,
, where:
A is the action name.
x1, , xn are individual variables, which denote the
objects on which the action operate.
PA is the set of pre-conditions, which must be satis-
fied before the action is executed, i.e. PA ={con|con
condition }.
EA is the set of post-conditions, which denote the ef-
fects of the action, EA is a set of pair headbody,
where head ={con | concondition}, body is a condi-
1) Action defines the transition relation of state, i.e. an
action A tra nsit a state u to a state v, if action A can pro-
duce state v under state u. The transition relation depends
on whether state u, v satisfy the pre-conditions and
post-conditions of action A. The transition relation is
denoted as u TA v.
2) The definition of condition and action description
reference [29,30].
Definition 3 (Parallel action stream) A parallel action
stream S is a n action stream that is achieved by the inde-
pendent relationship. S = a1a2ak, Which means the
candidate actions in the stream S can be independently
Definition 4 (Sequence action stream) S = a1; a2;;
ak. A sequence service S is a service that is achieved by
the prerequisite relationship.
Definition 5 (Action stream composition problem) the
composition problem can be described as a four-tuple <
T,A, G, S >, where:
T describes the vocabulary of the application domain.
A contains assertions about named individuals in
terms of this vocabulary and also denotes the initial
state of the world.
G is a set of assertions, which represent the goal at-
tempting to reach.
S is the set of action stream as described before.
The algorithm for finding the appropriate action stream
during the action composition is given in Algorithm 2.
Algorithm 2: Action composition
1) T is the Tbox
2) A is the ABox, add the initial state to A
3) G is the Goal
4) S = (S1,, Sn)
5) GoalStreams =
6) if T,A G then
7) GoalStreams = return GoalStreams
8) end if
9) select an assertion subGoal from the unsatisfied
goal G
Copyright © 2011 SciRes. IJIS
Z. Z. SHI ET AL.33
10) restGoal = G - subGoal
11) if the subGoal has already been processed by sub-
Stream then
12) executed recursively, and get restStreams to
reach restGoal.
13) GoalStreams = subStreamsjrestStreams
14) return GoalStreams
15) else if T,A subGoal then
16) executed recursively, and get restStreams to
reach restGoal.
17) GoalStreams = restStreams
18) return GoalStreams
19) else if If no action stream satisfy the subgoal sub-
Goal t then
20) return NULL
21) else
22) IF action stream S’ = {S’1,, S’k,} satisfying the
23) loop
24) select a set of streams subStreams from S’ sat-
isfying the subgoal
25) If a set of preconditions to satisfy the subgoal,
G’ = {G1, ···, Gm},
where Gi = Pre(subStreams)+
Posti(subStreams, subGoal)
26) loop
27) select a preGoal from G’ satisfying the sub-
28) executed recursively, and get preStreams to
reach preGoal.
29) executed recursively, and get restStreams to
reach restGoal.
30) if preStreams
NULL then
31) if restStreams
NULL then
32) GoalStreams = (preStreams; subStreams)
33) return GoalStreams
34) end if
35) end if
36) remove preGoal from G’
37) end loop
38) remove subStreams from S’
39) end loo p
40) return NULL
41) end if
Planning technologies differ in the complexity of the
problems they can handle and the representations. We
can employ different search algorithms to synthesis a
plans and the constraints they observe. A number of dif-
ferent planning methodologies are developed, such as
state space planning, plan space planning, planning graph
techniques, hierarchical task network planning, model
based planning. We will continue to consider which
method is better for action composition in the mind
model CAM.
4. Discussion and Conclusions
Cognitive cycle is a basic procedure of mental activities
in cognitive level. Cognitive cycle should reflect agent
interaction with environment, which is sensory outside
environment, through deliberation, and then affect it.
Traditional problem solving cycle can be embedded into
cognitive cycle and looks like a particular situation to
focus on solving questions. Here cognitive cycle is a
whole procedure from perception to behavior execution.
Franklin and Baars [15] pointed out “A cognitive cy-
cle can be thought of as a moment of cognition—a cog-
nitive moment; higher-level cognitive processes are
composed of many of these cognitive cycles, each a cog-
nitive atom.” And they proposed the cognitive cycle as
“perceive-understand-act”. This paper proposes the cog-
nitive cycle in CAM, which is different with LIDA’s
cognitive cycle. We emphasize motivation and action
composition. We use dynamic description logic to de-
scribe it formally.
We will continue to research on activities and algo-
rithms for the each phase in cognitive cycle. The pro-
posed cognitive cycle will be i mplemented by simulation
in multi-agent systems.
5. References
[1] A. Newell, “Unified Theories of Cognition,” Harvard,
Cambridge, 1990.
[2] J. Laird, A. Newell and P. Rosenbloom, “SOAR: An Ar-
chitecture for General Intelligence,” Artificial Intelligence,
Vol. 33, No. 1, 1987, p p. 1-6 4.
[3] J. R. Anderson, “The Adaptive Character of Thought,”
Erlbaum, Hillsdale, 1993.
[4] J. R. Anderson and C. Lebiere, “The Newell Test for a
Theory of Cognition,” Behavioral and Brain Science, Vol.
26, No. 5, 20 03 , pp. 601-648.
do i :1 0. 1017/ S 01 40525X0 300013X
[5] J. R. Anderson and Y. Qin, “Using Brain Imaging to Ex-
tract the Structure of Complex Events at the Rational
Time Band,” Journal of Cognitive Neuroscience, Vol. 20,
No. 9, 2008, pp. 1624- 1636. doi:10.1162/jocn.2008.20108
[6] J. R. Anderson, “Using Neural Imaging to Inform the
Instruction of Mathematics,” In: N. Zhong, K. C. Li, S. F.
Lu and L. Chen, Eds., Brain Informatics, International
Conference, 2009 International Conference on Brain In-
formatics, Beijing, 23-24 October, 2009, p. 1.
[7] J. R. Anderson, S. A. Betts, J. L. Ferris and J. M. Fin-
cham, “Neural Imaging to Track Mental States While
Using an Intelligent Tutoring System,” Proceedings of
the National Academy of Science, Vol. 107, No. 15 , 20 10 ,
Copyright © 2011 SciRes. IJIS
Copyright © 2011 SciRes. IJIS
pp. 7018- 70 23. doi:10.1073/pnas.1000942107
[8] Z. Z. Shi, “Advanced Artificial Intelligence,” World Sci-
entific Publishers, Hackensack, 2011.
[9] Y. Shoham, “Agent-Oriented Programming,” Artificial
Intelligence, Vol. 60, No. 1, 1993, pp. 51-92.
do i :1 0. 1016/ 00 04-3702(93) 90034-9
[10] S. Yoav and K. Leyton-Brown, “Multiagent Systems:
Algorithmic, Game-Theoretic, and Logical Foundations,”
Cambridge University Press, Cambridge, 2008.
[11] M. Georgeff, B. Pell, M. Pollack, M. Tambe and M.
Wooldridge, “The Belief-Desire-Intention Model of
Agency,” Intelligent Agents V. Agent Theories, Archi-
tectures, and Languages, 5th International Workshop,
ATAL’98, Pari s , 4-7 J uly 1998, pp . 630-6 30.
[12] R. Sun and X. Zhang, “Top-Down versus Bottom-up
Learning in Cognitive Skill Acquisition,” Cognitive Sys-
tems Research, Vol. 5, No. 1, 2004, pp. 63-69.
[13] R. Sun, “Motivational Representations within a Compu-
tational Cognitive A rc hitecture ,” 2008.
[14] S. Fran klin, B. J. Baars, U. Ramamurth y and M. Ventura,
“The Role of Consciousness in Memory,” Brains, Minds
and Media, No. 1, 2005 , pp. 1-3 8.
[15] B. J. Baars and S. Franklin, “Consciousness Is Computa-
tional: The LIDA Model of Global Workspace Theory,”
International Journal of Machine Consciousnesss, Vol. 1,
No. 1, 2009, p p. 2 3-32.
[16] T. Madl, B. J. Baars and S. Fr anklin, “The Timing of th e
Cognitive Cycle,” Plos One, Vol. 6, No. 4, 2011, Article
ID e14803. doi:10.1371/journal.pone.0014803
[17] Medical Dictionary Website, 2003.
[18] Z. Z. Shi and X. F. Wang “A Mind Model CAM in Intel-
ligence Science,” International Journal of Advanced In-
telligence, Vol. 3, No. 1, 2011, pp. 119-129.
[19] Z. Z. Shi, M. K. Dong, Y. C. Jiang and H. J. Zhang, “A
Logic Foundation for the Semantic Web,” Science in
China, Series F Information Sciences, Vol. 48, No. 2,
2005, pp. 161-178. doi:10.1360/03yf0506
[20] A. M. Collins and M. R. Quillian, “Retrieval Time from
Semantic Memory,” Journal of Verbal Learning and
Verbal Behavior, Vol. 8, No. 2, 1969, pp. 24 0- 2 47 .
[21] E. Tulving, “Elements of Episodic Memory,” Oxford
Clare ndon Press, Lon do n, 1983.
[22] A. Nuxoll and J. E. Larid, “Extending Cognitive Archi-
tecture with Episodic Memory,” Proceedings of the 22nd
AAAI Conference on Artificial Intelligence, Vol. 2, 20 07,
pp. 1560- 15 64.
[23] G. Tononi, G. M. Edelman and O. Sporns, “Complexity
and Coherency: Integrating Information in the Brain,”
Trends in Cognitive Sciences, Vol. 2, No. 12, 1998, pp.
474-484. doi:10.1016/S1364-6613(98)01259-5
[24] P. Langley, J. E. Laird and S. Rogers, “Cognitive Archi-
tectures: Research Issues and Challenges,” Cognitive Sys-
tems Research, Vol. 10, No. 2, 2009, pp. 141- 16 0.
[25] A. H. Maslow, “A Theory of Human Motivation,” Psy-
chol ogical Revi ew , Vol. 50, No. 4, 1943, pp. 370-396.
[26] A. H. Maslow, R. Frager and J. Fadiman, “Motivation
and Personality,” 3rd Edition, Addison-Wesley, Boston,
[27] J. Bach, “A Motivational System for Cognitive AI,”
AGI’11 Proceedings of the 4th International Conference
on Artificial General Intelligence, LNAI 6830, 2011, pp.
[28] J. Bach, “P SI—An Archi tectu re of Mo tivat ed Co gni tion,”
Oxford University Press, Oxford , 2009.
[29] F. Baader, M. Milicic, C. Lutz, U. Sattler and F. Wolter,
“Integrating Description Logics and Action Formalisms
for Reasoning about Web Services,” LTCS-Report
LTCS-05-02, Chair for Automata Theory, Institute for
Theoretical Computer Science, Dresden University of
Technology, Dresden, 2005.
[30] C. Lutz and U. Sattler, “A Proposal for Describing Ser-
vices with DLs,” In: I. Horrocks and S. Tessaris, Eds.,
Proceedings of the 2002 International Workshop on De-
scription Logics, Toulouse, 19-21 April 2002, pp. 129-