INTELLIGENT AGENT AND KNOWLEDGE MANAGEMENT
PERSPECTIVES FOR THE DEVELOPMENT OF INTELLIGENT
TUTORING SYSTEMS
Janis Grundspenkis
Department of Systems Theory and Design, Riga Technical University, 1 Kalku Street, LV 1658, Riga, Latvia
Keywords: Intelligent tutoring system, intelligent agent, multiagent system, knowledge management.
Abstract: The development of intelligent tutoring systems is discussed from intelligent agent and knowledge
management perspectives. A conceptual model in which both perspectives are integrated is proposed. The
model consists from system’s layer based on agent paradigm and knowledge worker’s layer responsible for
personal knowledge management of knowledge worker (teacher and/or student). The implemented
prototype of intelligent knowledge assessment system is described.
1 INTRODUCTION
Nowadays more and more modern organizations
realize that knowledge is their most important asset.
As a consequence a new type of intellectual work,
usually called a knowledge work, emerges. It is
obvious that teaching and learning also should be
changed to provide an effective turning of
information into knowledge. Concurrently one can
observe a rapid penetration of computer and
communication technologies into education that has
changed the traditional forms of teaching and
learning. Education from teacher-centered has
become student-centered (Waterhouse, 2004).
During last decades a lot of approaches, methods,
systems and environments has been proposed and
developed under the umbrella term of technology-
based learning.
Although today’s teaching and learning settings
are quite distinct from those of recent past, and more
distance education environments, e.g., eLearning,
mLearning, hybrid learning, etc. are used, the
experience shows that learning effectiveness is still
behind the desired level. The main reason is that the
intelligent support of teaching and learning
processes demonstrated by these systems is far
behind of that provided by the human teacher who is
able to adapt to each learner individually, to give a
flexible feedback (help, explanation, etc.) and to
assess the learner’s knowledge at all levels of the
well-known Bloom’s taxonomy (Bloom, 1956). All
mentioned issues (but not only) still are the
challenges for the developers of the intelligent
tutoring systems.
The first intelligent tutoring system SCHOLAR
(Carbonell, 1970) gave the origin to the successor
systems of such kind (BUGGY (Brown and Burton,
1978), GUIDON (Clancey, 1979), LISP Tutor
(Anderson and Reiser, 1985), ILIAD (Lincoln,
1991), ADIS (Warendorf and Tan, 1997), FLUTE
(Devedzic, Debenham and Popovic, 2000), BUT
(Butz, Hua and Maguire, 2004) are only some
examples). Faster progress can be observed when
Web-based intelligent tutoring systems become the
mainstream area of the research and development
(Yang, Kinshuk and Patel, 2002) and agent
technologies started to appear for the quality
improvement of Web-based education (Johnson,
2003).
The paper is organized as follows. In the second
section characteristics of intelligent tutoring system
(ITS) and its architecture are presented. The third
section is focused on applications of intelligent
agents in ITS. Knowledge management perspective
in ITS is discussed in the fourth section. A
conceptual model of ITS in which intelligent agent
and knowledge management perspectives are
integrated is proposed in the fifth section. The sixth
section gives the outline of implemented intelligent
knowledge assessment system which is part of ICT
under the development at the moment. Conclusions
380
Grundspenkis J. (2007).
INTELLIGENT AGENT AND KNOWLEDGE MANAGEMENT PERSPECTIVES FOR THE DEVELOPMENT OF INTELLIGENT TUTORING SYSTEMS.
In Proceedings of the Ninth International Conference on Enterprise Information Systems, pages 380-388
Copyright
c
 SciTePress

summarize the proposed approach and outline some
directions of future work.
2 CHARACTERISTICS AND
ARCHITECTURE OF
INTELLIGENT TUTORING
SYSTEMS
Regardless of variety of already developed methods
and systems the unambiguous definition of ITS is
not available. One definition which is mainly
focused on authoring systems is given in (Tennyson
and Christensen, 1988): “Intelligent tutoring systems
are inference-making systems that seek to
continuously improve the learning of each learner by
prescribing instruction that has a high probability of
preventing learner error or misconception, and by
continuously adapting this instruction according to
moment-to-moment diagnosis.” ITS’s use
knowledge about the domain, the student and about
teaching strategies to support flexible individualized
learning (Wenger, 1987; van Rosmalen and
Boticario, 2005).
The main characteristics of ITS are the
following:
 it is a computer based system,
 it uses methods of artificial intelligence such as
natural language processing, knowledge
representation, inference and machine learning
(Brusilovsky and Peylo, 2003),
 it is an adaptive system (Benyon and Murray,
1993),
 it simulates human teacher (supervisor),
 it tries to provide advantages of face-to-face
learning.
The ITS captures three types of knowledge:
 knowledge about “what to teach” (problem
domain knowledge),
 knowledge about “how to teach” (pedagogical
knowledge),
 knowledge about a learner (student).
Types of knowledge define an architecture of
ITS which consists from the expert module, the
pedagogical module, the student diagnosis module
and the interface module. The architecture of ITS is
shown in Figure 1.
It is worth to point out that the wide variety of
terms are used as synonyms to denote the
components of ITS. Examples are: expert module,
expert model, or expert solver; student diagnosis
module, student modeller, or user modeling
component; pedagogical module, curriculum and
instruction module, expert tutor, instruction model,
pedagogical model, tutorial module, or tutoring
component; interface, communication module,
graphical user interface shell, or supervisor unit, and
others.
Communication
module
Pedagogical
module
Expert
module
Student
diagnosis
module
Expert
model
Student
model
Pedagogical
model
Main components
Intelligent tutoring system
Additional components
Figure 1: The architecture of intelligent tutoring system.
INTELLIGENT AGENT AND KNOWLEDGE MANAGEMENT PERSPECTIVES FOR THE DEVELOPMENT OF
INTELLIGENT TUTORING SYSTEMS
381

Problem
To deliver
a problem
To receive
a solution
To give
a feedback
LEARNER
Solution
Feedback
Communication
module
To generate
a problem
Pedagogical
module
To receive
a solution
To work out
a feedback
Pedagogical
model
Problem
Solution
Feedback
Expert
module
To generate
a solution
Student
diagnosis
module
To get
the estimation
of current state
of student
knowledge
To compare
solutions
Mistake
library
Mistakes
Knowledge
assessment
Student’s
solution
Expert’s
solution
Student
model
Knowledge
assessment
Learning
goals
Student’s
knowledge
level
Figure 2: Operation schema of ITS.
To avoid misunderstanding in this paper the term
“module” is used to denote the identified component
of the system which have certain functionality while
the term “model” is used in the traditional meaning,
i.e., for simplified representation of an object. For
instance, expert model represents expert’s
knowledge, and expert module includes algorithms
for generation of problem solutions. Student model,
in its turn, represents information about each
particular student, and student diagnosis module
processes this information.
ITS may have also additional components, for
example, an explanation module to explain reasons
of mistakes or a development module to make
changes in the contents of the study course.
Communication module, i.e., an interface
provides functionality that allows to work with ITS.
The learner gets a problem from the pedagogical
module, gives the solution and receives a feedback.
Interactions between a learner and modules of ITS
are shown in Figure 2.
A pedagogical module provides the knowledge
infrastructure to adapt teaching and learning process
to needs and characteristics of each particular
learner. The main goal of this module is to decrease
and even to eliminate gap between the expert‘s
(teacher’s) and the student’s knowledge.
The role of a student diagnosis module is to
compare problem solutions given by a student and
an expert, to construct a student model and to use it
for estimation of current state of student knowledge.
The expert module is responsible for generation
of problem solutions which are passed to the student
diagnosis module for comparison.
3 AGENTS IN INTELLIGENT
TUTORING SYSTEM
A modern approach to artificial intelligence is
connected with the development and applications of
intelligent agents “that are anything that can be
viewed as perceiving its environment through
sensors and acting upon that environment through
effectors” (Russell and Norvig, 2003). An ITS can
be considered as a system of human agents
(supervisors and students) and/or software agents.
Software agents are programs that engage in dialogs,
and negotiate and coordinate transfer of information
(Murch and Johnson, 1999). Software agents
fundamentally differ from software packages
because they are user centered, autonomous, have
such attributes as adaptability, mobility,
transparency and accountability, ruggedness, self-
starter, social ability, reactivity, proactivity, and
learning capability (Grundspenkis and Anohina,
2004).
It is obvious that agents have many useful
features that are desirable for ITS. Agent perspective
provides several opportunities since the architecture
of ITS consists from several modules. So, each of
them can be implemented as an agent or a
multiagent system (Russell and Norvig, 2003). The
domain knowledge (study course content) may be
divided into knowledge units each of which can be
controlled by a separate agent. A pedagogical
module can provide different tutoring strategies and
each of them can be entrusted to the separate agent.
Information stored in a student model can be
ICEIS 2007 - International Conference on Enterprise Information Systems
382

categorized into several classes: learning styles,
psychological characteristics, causes of mistakes and
misconceptions, etc. It is possible to develop an
agent that will be responsible for gathering and
processing the certain type of information. ITS can
support various interaction devices that allow the
learner to communicate with the system. In this case
it is possible to develop agents responsible for
management of different devices, such as, for
instance, monitor + mouse + keyboard or data glove
+ motion tracking + voice recognition. The same
approach can be applied for user interface which
includes various tools, e.g. buttons, menus, input
fields, panels, and so on.
In fact, all components of ITS described in the
previous section can be implemented as agents.
Analysis of already developed ITS, namely, Ines
(Hospers, et.al., 2003), ABITS (Capuano, et.al.,
2000), WADIES (Georgouli, Paraskakis and
Guerreiro, 2003), IVTE (Nunes, et.al., 2002), a
multi-agent architecture for distance education
systems (Dorea, Lopes and Fernandes, 2003) and
intelligent virtual environment for training (De
Antonio, et.al., 2003), allows to outline the possible
agent-based solutions for all ITS modules
(Grundspenkis and Anohina, 2004).
Agents in the pedagogical module can evaluate,
update and generate curriculum, implement different
teaching strategies (a case when a multiagent
architecture is required), and generate a feedback
(explain and to provide help). So, the typical set of
agents in a pedagogical module is a curriculum
agent, a feedback and explanation agent, and
teaching strategy agents (one for each available
teaching strategy). The main task of agents that
comprise a student diagnosis module is the
evaluation and updating of information about a
particular learner. In this case agent functions are
building a profile of learner’s psychological
characteristics (learning preferences, learning style,
attentiveness, etc.), building a model of learner’s
current state of knowledge and skills, registering
learner’s mistakes and his/her history of interactions
with the system.
The set of typical student modelling agents of
student diagnosis module includes cognitive
diagnosis agent, psychological agent, knowledge
evaluation agent, interaction registering agent, and
mistake registering agent. The set of agents in an
expert module strongly depends on the problem
domain and, as a rule, multiagent architecture is
used in this module. The number of agents used
depends on the number of knowledge units in which
a study course content is decomposed. If a
communication module is based on agent
technologies, agent functions are management of
different interaction tools and devices, and
monitoring the interaction between the learner and
the system.
A typical set of agents which may constitute the
architecture of ITS is shown in Figure 3.
Of course, ITS can contain also specific agents,
that aren’t included in typical sets of agents, and are
determined by specific features of the problem
domain or peculiarity of ITS architecture and
technologies used for its implementation. Several
examples of specific agents may be found, for
instance, the authoring agent in WADIES, or the
spooler agent in multiagent architecture of ABITS. It
is worth to point out that in several ITS animated
pedagogical agents are used. Such agent emulates
aspects of dialogue between a human teacher and a
learner.
Communication
module
implemented
as a set of agents
Teaching
strategy 1
agent
Teaching
strategy N
agent
...
Curriculum
agent
Feedback
agent
Pedagogical module
Cognitive
diagnosis
agent
Knowledge
evaluation
agent
Psychological
agent
Interaction
registering
agent
Mistake
registering
agent
Student diagnosis module
Expert
agent 1
Expert
agent M
...
Expert module
INTELLIGENT TUTORING SYSTEM
Figure 3: A typical set of agents included in ITS.
INTELLIGENT AGENT AND KNOWLEDGE MANAGEMENT PERSPECTIVES FOR THE DEVELOPMENT OF
INTELLIGENT TUTORING SYSTEMS
383

In accordance with (Baylor and Kim, 2003)
animated pedagogical agents can play three roles:
agent as an expert, agent as a motivator and agent as
a mentor. Some details about animated pedagogical
agents (including widely known Steve and Adele)
also called guidebots can be found in (Johnson,
2003).
4 KNOWLEDGE MANAGEMENT
PERSPECTIVE IN
INTELLIGENT TUTORING
SYSTEM
In all education systems regardless of their kind
(face-to-face, distance, mobile, hybrid, etc.) there are
two groups of actors, namely, supervisors (teachers)
and students (learners) who are working with
knowledge. In (Grundspenkis, 2005) a conceptual
model is proposed in which actors of the intelligent
tutoring system are considered to be the knowledge
workers embedded into a knowledge management
system (KMS) as it is shown in Figure 4.
Group of students
Supervisor
Intelligent tutoring system
Knowledge management system
Figure 4: Intelligent tutoring system embedded in its
environment.
A KMS is an infrastructure of mutually integrated
techniques and tools created for such knowledge
management process support as knowledge
acquisition, processing, distribution and usage, as
well as for generation of new knowledge. Following
(Thierauf, 1999) the main KMS functions are:
 detection of information and/or knowledge,
 storage of information and/or knowledge,
 inference of conclusions,
 retrieval and visualization of knowledge,
 decision making.
The KMS enables to turn information into action
and to connect people to knowledge, i.e., enables an
effective and active learning process. The most
important aspects of effective learning process are
construction of knowledge, co-operation and
teamwork, and learning through problem solving. In
other words, the KMS supports expansion of
individual’s personal knowledge to the knowledge
of a group as a whole. It means that knowledge
management environment must contribute both
personal knowledge and organizational knowledge
as well. In this context a concept of personal
knowledge management (PKM) emerges. PKM is a
collection of processes that an individual needs to
carry out in order to gather, classify, store, search
and retrieve knowledge in his/her activities (Tsui,
2002). PKM is an integrative discipline that
integrates many aspects and many perspectives from
different fields. A PKM system (PKMS) is a
complex system that includes psychological, social
and technological aspects: individual’s emotional
intelligence, his/her understanding and aims,
environment and society where he/she lives in and
acts, as well as technologies (Apshvalka and
Grundspenkis, 2005). It is quite obvious that
practically all mentioned aspects are important in
teaching and learning process.
Let have a closer look on why knowledge
management may play an important role in ITS
development. First, each educational organization
must enhance its knowledge assets or at least must
keep them on the needed level. Unfortunately,
education organizations very easily may lose their
knowledge assets when teachers are leaving. To
avoid loses (at least to the certain extent) educational
organization must extend its intellectual capital.
According to (Stewart, 1994) an intellectual capital
is an intellectual material that has been formalized in
some useful order, captured in a way that allows it to
be described, shared, distributed, and leveraged to
produce a higher valued asset. So, it may be
effectively supported by the KMS.
Different types of knowledge that education
organization possesses and various knowledge
possessors in it is the second factor why knowledge
management may play an important role in the
context of ITS. The widely known classification of
knowledge into two classes, namely, tacit
knowledge and explicit knowledge is proposed in
(Nonaka and Takeuchi, 1995). Tacit knowledge is
personal knowledge embedded in individual
experience. Most commonly it is shared and
exchanged through direct, face-to-face contact and
can be communicated in a direct and effective way.
Nowadays technologies help and it is not necessary
to store all needed knowledge in human brains. In
many cases including education it is enough to know
ICEIS 2007 - International Conference on Enterprise Information Systems
384

where to find necessary information and knowledge
and to be able to get it quickly enough. This is where
PKMS should help.
The third aspect showing the potential role of
knowledge management in ITS development is the
mode in which collection and retrieval of knowledge
is performed. Details of this aspect which is closely
connected with notion of corporate memory are
beyond the scope of this paper and can be found in
(van Heijst, van der Spek and Kruizinga, 1998).
5 CONCEPTUAL MODEL OF
INTELLIGENT TUTORING
SYSTEM
Both discussed perspectives, that is, agent and
knowledge management perspectives are integrated
in a conceptual model of ITS. The ITS itself is under
the development. The conceptual model has two
layers – system’s layer and knowledge worker’s
layer. Functioning of both layers is supported by sets
of agents. At the system’s layer all components of
ITS shown in Figure 1 are included. The knowledge
worker’s layer supports students and supervisors
involved in the teaching and learning process.
Students and supervisors are knowledge workers,
i.e., human agents which are working with
knowledge. To make their work effective the PKMS
is used which is based on a set of agents operating at
the working place of each knowledge worker.
Agents are located in three circles (Grundspenkis
and Kirikova, 2005) as it is shown in Figure 5.
Knowledge worker
Inner circle
Personal agents
Medium circle
Internal communication agents
External circle
External communication agents
Figure 5: Agents of the knowledge worker.
Personal agents are search, assistant, filtering and
workflow agents (Knapik and Johnson, 1998).
Search agents are used to search titles of documents
or directories on the Web. Filtering agents may
monitor the data stream searching the text for
knowledge and phrases as well as the list of
synonyms, and try to forward only the useful
information. Assistant agents usually are designed to
wait for events such as E-mail messages to occur
and then to sort them by sender, subject, time,
priority, etc. Workflow agents are useful for task
coordination and meeting scheduling. Smart agents
(Case, et.al., 2001) may appear in near future that
will be able to acquire, store, generate and distribute
knowledge.
Internal communication agents provide
communications between individuals. This set of
agents includes messaging, team, collaborative and
cooperative agents. Messaging agents can connect
students within a group and with the supervisor no
matter where they are and what communication
medium is used. Team agents facilitate
communication in the group of students, while
cooperative and collaborative agents are able to
cooperate and to collaborate with filtering agents in
the internal circle.
Agents for communication with external systems
are, for instance, network agents, database agents,
connection and access agents, and intelligent Web
agents. It is obvious that from the knowledge
management perspective of ITS the most important
role may play intelligent Web agents because
nowadays the Web is the richest source of data,
information and knowledge that is useful for
learning and is accessible for any user. At the same
time currently the Web contains a lot of data,
structured data (structured documents, online
databases), simple metadata but very little
knowledge, i.e., very few formal knowledge
representations (Web intelligence, 2003). The reason
is that the knowledge is encoded using various
languages and practically unconnected ontologies.
As a consequence, each knowledge source requires
the development of special wrapper for its
knowledge to be interpreted and hence retrieved,
combined and used. Efforts to solve this problem
resulted in the appearance of a new paradigm, so
called Web intelligence (Web intelligence, 2003)
which is challenging and promising research field
for ITS developers.
6 IMPLEMENTATION OF
INTELLIGENT KNOWLEDGE
ASSESSMENT SYSTEM
At the present moment part of the system’s layer of
conceptual model already has been developed,
implemented and tested. Due to the scope of this
INTELLIGENT AGENT AND KNOWLEDGE MANAGEMENT PERSPECTIVES FOR THE DEVELOPMENT OF
INTELLIGENT TUTORING SYSTEMS
385

paper in this section only the outline of the
developed prototype is given details of which may
be found in (Anohina and Grundspenkis, 2006).
In process oriented learning a teacher divides a
study course into several stages. At the end of each
stage the teacher gives assessment of learner’s
knowledge level. Assessment is based on the notion
of concept maps. Concept maps are a special kind of
mental model and a method for representation and
measuring of individual’s knowledge (Croasdell,
Freeman and Urbaczewski, 2003). Concept maps are
represented by graphs. Nodes represent concepts and
arcs represent relationships between concepts.
The system’s architecture includes modules of
administrator, teacher and learner. The modules
interact using a common database. The administrator
maintains the system and manages data about
individual learners, groups of learners, as well as
teachers and the courses. The teacher’s module
supports the development of concept maps and
examination of learner-completed concept maps.
The learner’s module includes tools for completion
of concept maps and viewing the feedback after the
evaluation of the correctness of learner’s concept
maps.
The core of the developed system is the
intelligent assessment agent which is shown in
Figure 6.
The communication agent perceives the learner’s
actions, i.e. concepts inserting into and removing
from the structure of a concept map. It is also
responsible for visualization of a structure of a
concept map received from the agent-expert, and for
the output of feedback received from the knowledge
evaluation agent. After the learner has confirmed
his/her solution, the communication agent delivers
the learner-completed concept map to the knowledge
evaluation agent which compares the concept maps
of the learner and the teacher and recognizes five
patterns of learner’s solution (correct and incorrect
ones).
The interaction registering agent receives the
learner-completed concept map from the
communication agent and results of its comparison
from the knowledge evaluation agent, and stores
them in a database. The agent-expert forms a
structure of a concept map of the current stage on
the basis of the teacher-created concept map and the
learner’s concept map of previous stage. The formed
structure is passed to the communication agent for
its visualization. The agent-expert also delivers a
teacher-created concept map to the knowledge
evaluation agent for comparison. The prototype has
been developed using the following tools: Borland
JBuilder 9.0, JGraph, PostgreSQL DBMS 8.0.3 and
JDBC drivers for PostgreSQL.
The operation of the developed system has been
tested in four study courses (two engineering courses
“Systems Theory Methods” and “Fundamentals of
SQL”, and two courses of social sciences). Seventy
four students have been involved in testing, 57% of
them found that completing of concept maps was
difficult for them, and 33% found it easy.
Database
of teacher’s
concept maps
Database
of learner’s
concept maps
and assessments
Communication agent
A
gent-
expert
Interaction
registering
agent
Knowledge
evaluation
agent
INTELLIGENT ASSESSMENT AGENT
Teacher-created
concept map
Structure
of the concept map
of the current stage
Learne
r
-
completed
concept map
Feedback
LEARNER
Structure
of the concept map
of the current stage
Learne
r
-
completed
concept map
Feedback
Learne
r
-
completed
concept map
Assessment
Learner’s
concept map
and assessment
Learner’s
concept map
of the previous
stage
Teache
r
-created
concept map
Figure 6: The architecture of the intelligent assessment agent.
ICEIS 2007 - International Conference on Enterprise Information Systems
386

More than a half (62%) answered that work with
the developed system helped them to understand
course contents better.
7 CONCLUSIONS
In this paper a system approach is used to integrate
intelligent agent and knowledge management
perspectives for the development of intelligent
tutoring systems. The tutoring system is based on the
intelligent agent paradigm and is embedded in the
knowledge management system which plays the role
of its environment. A synergy effect is expected
from such kind of integration especially in hybrid
course development where a part of contents is
taught in the traditional face-to-face manner, and
another part using distance learning facilities.
The implementation of the proposed conceptual
model of intelligent tutoring system which consists
from system’s layer and knowledge worker’s layer is
already started. The prototype of the intelligent
knowledge assessment system based on concept
maps has been developed and tested. The prototype
modules partly cover modules of traditional
architecture of intelligent tutoring systems:
implemented communication agent is one of needed
agents of communication module; agent-expert
realizes an expert module; knowledge assessment
agent and interaction registering agent are part of
student diagnosis module.
Despite that fact that a lot of work is needed to
implement the proposed conceptual model as a
whole, the developed prototype has a good potential
for further evolution and research. One of planned
directions of future work is to use ontologies for
more flexible knowledge assessment taking into
account semantics of links. To achieve this goal it is
needed to develop algorithms and tools for concept
map generation from course ontology, and
algorithms for concept map evaluation. It is also
necessary to improve feedback given to the teacher
and to the learner. For the later the system should
generate recommendations related to the learning
material that the learner should revise to fill gaps in
his/her knowledge.
REFERENCES
Anderson, J.R., Reiser, B.J., 1985. The LISP Tutor. Byte,
Vol. 10, 159-175.
Anohina, A., Grundspenkis, J., 2006. Prototype of
multiagent knowledge assessment system for support
of process oriented learning. In Proceedings of the 7th
International Baltic Conference on Databases and
Information Systems, July 3-6, 2006, Vilnius,
Lithuania, IEEE Operation Center, Lithuania, 211-
219.
Apshvalka, D., Grundspenkis, J., 2005. Personal
knowledge management and intelligent agent
perspective. In Proceedings of the 14th International
Conference on Information Systems Development Pre-
Conference-ISD 2005, Karlstad, Sweden, 14-17
August, 2005, Nilsson, A.G. et.al. (eds.), Karlstad
University Studies, Karlstad, Sweden, 219-230.
Baylor, A.L., Kim, Y., 2003. Validating pedagogical agent
roles: expert, motivator, and mentor. In Proceedings of
EDMedia, 15th Annual Conference on Educational
Multimedia, Hypermedia, and Telecommunications,
Honolulu, Hawaii, June 2003, 463-466.
Benyon, D.R., Murray, D.M., 1993. Adaptive systems:
from intelligent tutoring to autonomous agents.
Knowledge-Based Systems, December 1993, Vol. 6,
No. 4, 197-219.
Bloom, B.S., 1956. Taxonomy of education objectives.
Handbook I: the cognitive domain. David McKay Co
Inc., New York.
Brown, J.S., Burton, R.R., 1978. A paradigmatic example
of an artificially intelligent instructional system.
International Journal of Man-Machine Studies, Vol.
10, 323-339.
Brusilovsky, P., Peylo, C., 2003. Adaptive and intelligent
web-based educational systems. International Journal
of Artificial Intelligence in Education, No. 13, 156-
169.
Butz, C.J., Hua, S., Maguire, R.B., 2004. A web-based
intelligent tutoring system for computer programming.
In Proceedings of the IEEE/WIC/ACM International
Conference on Web Intelligence (WI’04), Beijing,
China, September 20-24, 2004, 159-165.
Capuano, N., De Santo, M., Marsella, M., Molinara, M.,
Salerno, S., 2000. A multi-agent architecture for
intelligent tutoring. In Proceedings of the
International Conference on Advances in
Infrastructure for Electronic Business, Science, and
Education on the Internet, SSGRR 2000, L’Aquila,
July 31-August 06, 2000.
http://www.capuano.biz/Papers/
SSGRR_2000_P1.pdf (visited 02.10.2004)
Carbonell, J.R., 1970. AI in CAI, An artificial intelligence
approach to computer-assisted instruction. IEEE
Transactions on Man-machine Systems, 11(4), 190-
202.
Case, S., et.al., 2001. Enhancing e-communities with
agent-based systems. Computer, July 2001, 64-69.
Clancey, W.J., 1979. Tutoring rules for guiding a case
methods dialogue. In Intelligent Tutoring Systems,
Sleeman, D.H., Brown, J.S. (eds.), Academic Press,
New York, 201-225.
Croasdell, D.T., Freeman, L.A., Urbaczewski, A., 2003.
Concept maps for teaching and assessment.
Communications of the Association for Information
Systems, Vol. 12, 396-405.
INTELLIGENT AGENT AND KNOWLEDGE MANAGEMENT PERSPECTIVES FOR THE DEVELOPMENT OF
INTELLIGENT TUTORING SYSTEMS
387

De Antonio, A., Imbert, R., Ramirez, J., Mendez, G.,
2003. An agent-based architecture for the development
of intelligent virtual training environments. In
Proceedings of the Second International Conference
on Multimedia and Information & Communication
Technologies in Education (m-ICTE 2003), Badajoz,
Espana, 3-6 de Diciembre, 2003, 1944-1949.
Devedzic, V., Debenham, J.K., Popovic, D., 2000.
Teaching formal languages by an intelligent tutoring
system. Educational Technology & Society, Vol. 3,
No. 2. http://ifets.ieee.org/periodical/vol_2_2000/
devedzic.html (visited 18.04.2005)
Dorea, F.A., Lopes, C.R., Fernandes, M.A., 2003. A
multiagent architecture for distance education systems.
In Proceedings of the 3rd IEEE International
Conference on Advanced Learning Technologies
(ICALT’03), Athens, Greece, July 9-11, 2003, 368-
369.
http://csdl.computer.org/comp/proceedings/icalt/2003/
1967/00/19670368.pdf (visited 06.10.2004)
Georgouli, K., Paraskakis, I., Guerreiro, P., 2003. A web
based tutoring system for compilers. In Proceedings of
the 14th EAEEIE Annual Conference on Innovation in
Education for Electrical and Information Engineering
(EIE), Gdansk, Poland, 16th - 18th June, 2003.
http://www-ctp.di.fct.unl.pt/~pg/docs/
Gdansk_2003_2.pdf (visited 02.10.2004)
Grundspenkis, J., Anohina, A., 2004. Agents in intelligent
tutoring systems: state of the art. In Scientific
Proceedings of Riga Technical University, 5th series,
Computer Science, Applied Computer Systems, Vol.
22, RTU Publishing, Riga, Latvia (2004), 110-121.
Grundspenkis, J., 2005. Conceptual framework for
integration of multiagent and knowledge management
techniques in intelligent tutoring systems. In
Information Systems Development. Advances in
Theory, Practice, and Education. Vasilecas, O., et.al.
(eds.), Springer Science + Business Media, 207-216.
Grundspenkis, J., Kirikova, M., 2005. Impact of intelligent
agent paradigm on knowledge management. In
Intelligent Knowledge-Based Systems, Vol. 1,
Leondes, C.T. (ed.), Kluwer Academic Press, Boston,
Dordrecht, London, 164-206.
Hospers, M., Kroezen, L., Nijholt, A., Akker, R., Heylen,
D., 2003. Developing a generic agent-based intelligent
tutoring system and applying it to nurse education.
http://wwwhome.cs.utwente.nl/~heylen/Publicaties/
ines-2003.pdf (visited 13.08.2004)
Johnson, W.L., 2003. Using agent technology to improve
the quality of web-based education. In Web
Intelligence, Zhong, N., Liu, J., Yao, Y.Y. (eds.),
Springer-Verlag, Berlin, Heidelberg, 77-101.
Knapik, M., Johnson, J., 1998. Developing intelligent
agents for distributed systems. McGraw Hill, NY.
Lincoln, N., 1991. ILIAD training enhances students’
Diagnostic Skills. Journal of Medical Systems, Vol.
15, 93-110.
Murch, R., Johnson, T., 1999. Intelligent software agents.
Prentice Hall, Upper Saddle River, N.J.
Nonaka, I., Takeuchi, H., 1995. Knowledge creating
organizations. Oxford University Press, New York.
Nunes, M.A.S., Dihl, L.L., Fraga, L., Woszezenki, C.R.
et.al., 2002. Multi-agent systems applied to intelligent
tutoring systems. International Conference on
Artificial Intelligence IC-AI’02, Las Vegas, Nevada,
USA, June 24-27, 2002, 361-366.
http://www.inf.ufsc.br/~lisi/Lasvegas.pdf (visited
06.10.2003)
Russell, S.J., Norvig, P., 2003.
Artificial intelligence. A
modern approach. Prentice Hall, Upper Saddle River,
N.J.
Stewart, T.A., 1994. Your company’s most valuable asset:
intellectual capital, Fortune, 130(68), 68-74.
Tennyson, R.D., Christensen, D.L., 1988. MAIS: an
intelligent learning system. In Instructional design for
microcomputer courseware, Jonassen, D.H. (ed.),
Erlbaum, Hillsdale, NJ, 247-274.
Thierauf, R.J., 1999. Knowledge management systems for
business. Quorum Books, Westport, Connecticut,
London.
Tsui, E., 2002. Technologies for personal and peer-to-peer
(P2P) knowledge management, CSC Leading Edge
Forum Technology Grant report.
van Heijst, G., van der Spek, R., Kruizinga, E., 1998. The
lessons learned cycle. In Information Technology of
Knowledge Management, Borghoff, U.M., Pareschi,
R. (eds.), Springer Verlag, Berlin, Heidelberg, New
York, 17-34.
van Rosmalen, P., Boticario, J., 2005. Using learning
design to support design and runtime adaptation. In
Learning Design, Koper, R., Tattersall, C. (eds.),
Springer-Verlag, Berlin, Heidelberg, 291-301.
Warendorf, K., Tan, C., 1997. ADIS – an animated data
structure intelligent tutoring system or putting an
interactive tutor on the WWW. In Proceedings of the
Workshop “Intelligent Educational Systems on the
World Wide Web” at the 8th World Conference of the
AIED Society, Kobe, Japan, 18-22 August 1997, 54-
60. http://www.contrib.andrew.cmu.edu/~plb/
AIED97_workshop/Warendorf/Warendorf.html
(visited 18.04.2005)
Waterhouse, S., 2004. The power of eLearning: the
essential guide for teaching in the digital age. Allyn &
Bacon.
Wenger, E., 1987. Artificial intelligence and tutoring
systems. Morgan Kaufman, San Francisko.
Yang, A., Kinshuk, Patel, A., 2002. A plug-able web-
based intelligent tutoring system. In Proceedings of
the Xth European Conference on Information Systems,
ECIS 2002, Vol. 2, Wrycza, S. (ed.), Wydawnictwo
Uniwersytetu Gdanskiego, Gdansk, Poland, 1422-
1429.
Web intelligence, 2003. Zhong, N., Liu, J., Yao, Y.Y.
(eds.), Springer-Verlag, Berlin, Heidelberg.
ICEIS 2007 - International Conference on Enterprise Information Systems
388