Techniques, Technologies and Patents Related to Intelligent Educational Systems - E-Infrastructures and Technologies for Lifelong Learning - page 5

Figure 1. Basic architecture of an intelligent educational system

BACKGROUND

basic components. We discuss issues and problems
arising in IESs and present relevant solutions.
Solutions come from AI methods and patents
related to IESs. We outline functions that can be
implemented by AI methods in different phases
of an IES lifecycle. We survey several recent
patents related to IESs categorizing them accord-
ing to various aspects. At the end, we specify
future trends regarding IESs. This chapter may
be used as a guide by researchers, students and
professionals working in the field of e-learning
and lifelong learning as it presents state-of-the-art
technologies/techniques in IESs, relevant work
that has been patented and future directions. To
the authors' best knowledge, no survey related to
AI methods/technologies suitable for IESs and
categorization of patents related to IESs has been
presented till now. (Hatzilygeroudis & Prentzas,
2009) presents an initial survey on recent patents
related to IESs.
The structure of the chapter is as follows. The
following section describes the necessary back-
ground knowledge. The third section presents
functions and issues regarding IES modules as
well as corresponding solutions from AI methods
and patents related to IESs. Then the fourth section
discusses issues regarding future research direc-
tions for IESs. Finally, conclusions are drawn.

The first computer-based educational systems
were called Computer-Assisted Instruction
(CAI) systems. CAI systems started in the 1950s
as simple 'linear programs', which evolved
afterwards. In the 1960s, 'branching programs'
offered corrective feedback adapting teaching
to learner responses. In the 1970s, 'generative'
systems appeared that in certain domains (e.g.
arithmetic) could generate learning content them-
selves (Yazdani, 1988). A major disadvantage of
those systems is their inability to adapt instruc-
tion to learners' diversity and individual needs.
Evaluation of learner knowledge is not performed
intelligently, but based on final responses to 'yes-
no' and multiple-choice questions. In addition,
sequencing strategies of learning items in CAIs
follow traditional linear and branching approaches
(Brusilovsky & Peylo, 2003). So, CAI systems
cannot answer learner questions concerning 'why'
and 'how' the task is performed (Yazdani, 1988).
IESs surpass such drawbacks.
Intelligent Tutoring Systems (ITSs) constitute a
popular type of IESs. ITSs take into account learner
knowledge level and skills and adapt learning
content presentation to needs and abilities of him/
her (Polson & Richardson, 1988; Yazdani, 1988;
Woolf, 1992). ITSs traditionally lay emphasis
on AI techniques to achieve their tasks. An ITS
should be able to perform tutoring tasks such as