Information Technology Reference
In-Depth Information
1.2.4 Machine Learning Techniques
The student model is responsible for the identification of the student's knowledge
level, misconceptions, needs and preferences. This kind of information is obtained
by observing the student's behavior and action during her/his interaction with the
adaptive and/or personalized tutoring system. The processes of the student's behavior
observation and reasoning should be made automated by the system. This is achieved
by machine learning techniques. Machine learning concerns the formation and
study of models that allow the system to learn from observation's data and make
automatically inferences (Webb 1998). Machine learning have so far been used
either to induce a single, consistent student model from multiple observed stu-
dent behaviors, or for the purpose of automatically extending or constructing from
scratch the bug library of student modelers (Sison and Shimura 1998). Therefore,
machine-learning techniques can be used to predict future actions (Webb et al.
2001) and make the system able to adapt the instruction and learning processes to
the student's needs. An approach of machine learning is the use of artificial neural
networks. They are computational systems inspired by the biological nervous system
of the brain. Artificial neural networks are presented as interconnected networks of
“neurons” that can learn through experience via algorithms.
1.2.5 Cognitive Theories
The adaptive and/or personalized tutoring systems have to integrate pedagogical
and psychological theories, except of artificial intelligence, to be effective. Indeed,
many researchers (e.g. Salomon 1990; Welch and Brownell 2000) have pointed
out that technology is effective when developers thoughtfully consider the merit
and limitations of a particular application while employing effective pedagogi-
cal practices to achieve a specific objective. Pedagogical practices can be inte-
grated in a student model by using cognitive theories, which attempt to explain
human behavior during the learning process. Cognitive theories can model either
the student's cognitive characteristics like knowledge, attention, ability to learn
and understand and memory or the student's emotional states and motivation.
Therefore, they contribute significantly to the student's reasoning trying to under-
stand human's processes of thinking and understanding.
There are a variety of cognitive theories. Some cognitive theories that have
been used in student modeling are: the Human Plausible Reasoning (HPR)
theory (Collins and Michalski 1989), which is a domain-independent theory that
categorizes plausible inferences in terms of a set of frequently recurring infer-
ence patterns and a set of transformations on those patterns (Burstein and Collins
1988; Burstein et al. 1991); the Ortony et al. (1988) (OCC) theory, which allows
modeling possible emotional states of students, and the Control-Value theory
(Pekrun et al. 2007), which is an integrative framework that employs diverse
factors, e.g. cognitive, motivational and psychological, to determine the existence
Search WWH ::




Custom Search