Techniques, Technologies and Patents Related to Intelligent Educational Systems - E-Infrastructures and Technologies for Lifelong Learning - page 6

selecting/sequencing of presented learning items,
responding to learner questions concerning in-
structional goals and content, in-depth analysis of
learner responses to presented problems/questions
and determining when assistance is needed and
how to provide it (Polson & Richardson, 1988).
ITSs are suitable for lifelong learners as they can
provide the most suitable learning activities to
meet learner goals.
Another type of educational systems providing
personalization is Adaptive Educational Hyper-
media Systems (AEHSs) (Brusilovsky, Kobsa &
Vassileva, 1998). AEHSs are specifically devel-
oped for hypertext environments, such as the Web.
They use technologies/techniques from Adaptive
Hypermedia (AH). Main services offered are adap-
tive presentation of learning content and adaptive
navigation by adapting page hyperlinks. Compared
to ITSs, they offer greater sense of freedom to
learners, since they enable a guided navigation
to user-adapted educational pages. Furthermore,
they dynamically construct or adapt educational
pages whereas in ITSs educational pages contents
are typically static (Papanikolaou et al., 2003). A
number of AHESs use efficient but simple tech-
niques that can hardly be considered as 'intelligent'
(Brusilovsky & Peylo, 2003). Enhancing AEHSs
with ITS techniques creates another type of IESs.
AEHSs enhance self-directed lifelong learning as
they provide advice/guidance to identify the most
suitable learning activities matching their needs.
The fields of ITSs and AEHSs were well
established before the Internet age (Brusilovsky
& Peylo, 2003). During that period, ITSs and
AEHSs were usually developed as stand-alone
systems. However, emergence of the Web gave
rise to various Web-based ITSs and AEHSs,
called Web-Based Intelligent Educational Sys-
tems (WBIESs) (Hatzilygeroudis, 2004). Several
of them combine ITS and AEHS technologies.
WBIESs are accessible by a large number of learn-
ers giving the opportunity for effective lifelong
learning experiences as well as thorough testing
and subsequent refinement of their mechanisms/

services. Large quantities of learner data are
stored in WBIESs that can be exploited to extract
knowledge regarding the pros and cons of IES
learning process. Therefore, the Web offers IESs
the chance to become more acceptable by those
seeking lifelong learning. As the Web imposes
constraints on learning process, the role of AI
methods is becoming increasingly important.

CHALLENGES IN
DEVELOPING INTELLIGENT
EDUCATIONAL SYSTEMS

In this section, we discuss issues/problems regard-
ing IESs and present certain solutions/recom-
mendations. Issues involve all three primary IES
modules (domain knowledge, user modeling unit,
pedagogical module), all stages of IES life cycle
(construction, operation and maintenance) and all
user types (authors, tutors, learners, knowledge
engineers). We also present brief background
information concerning IES module functionality.

Domain Knowledge Issues

Domain knowledge contains knowledge regarding
the learning subject and actual learning content. It
usually consists of two parts: knowledge concepts
and learning units . Knowledge concepts refer to
basic domain knowledge entities. Relations can
be defined among knowledge concepts, among
learning units as well as among concepts and
learning units. Typical relations among concepts
are the following:

• Prerequisite : Some concepts are prerequi-
site of others. A learner should know some
or all prerequisite concepts of a concept to
proceed to it.
• Part-of : Many simpler concepts are part of
a more complex concept.
• Is-a : This relation connects a concept with
others that are its typical instances.