Information Technology Reference
In-Depth Information
Web-based educational systems offer easy access to knowledge domains
and learning processes from everywhere for everybody at any time. As a result,
users of web-based educational systems are of varying backgrounds. They have
heterogeneous needs, different levels of knowledge and abilities. That is the
reason why researches in the field of e-learning have expanded their interests on
adaptive e-learning, which is suitable for teaching heterogeneous student popu-
lations (Schiaffino et al. 2008). An adaptive system must be capable of managing
learning paths adapted to each user, monitoring user activities, interpreting those
using specific models, inferring user needs and preferences and exploiting user
and knowledge domain to dynamically facilitate the learning process (Boticario
et al. 2005). In other words, an adaptive educational system has to provide
personalization to the specific needs, knowledge and background of each indi-
vidual student.
A solution is the student model. Student modeling has been introduced in
Intelligent Tutoring Systems, but its use has been extended to most current
educational software applications that aim to be adaptive and personalized.
A student model allows understanding and identification of student needs.
By keeping a model for every user, a system can successfully personalize its
content and utilize available resources accordingly (Kyriacou 2008). For
example, in an adaptive educational application, a student model can be used
to achieve accurate student diagnosis and predict a student's needs. In return,
it offers individualized courses (Gaudioso et al. 2010), adaptive navigation
support (Castillo et al. 2009), help and feedback to students (Tsiriga and Virvou
2003a; Chrysafiadi and Virvou 2008), allowing them to learn in their own pace
(Chrysafiadi and Virvou 2013c).
The student's model dimensions and properties correspond to the physical
student's features and characteristics (Yang et al. 2010). Therefore, in order to
construct a student model, it has to be considered what information and data about
a student should be gathered. The student's characteristics are: the knowledge
level, the errors and misconceptions, the learning preferences and style, other cog-
nitive features, the emotions, the motivation and meta-cognitive skills. To model
them there is a variety of student modeling techniques to choose: overlay model,
stereotypes, perturbation, constraint-based model, learning machine algorithms,
fuzzy logic, Bayesian networks etc. (Chrysafiadi and Virvou 2013b). However in
most cases there is the need to model more than one student's characteristics. That
is achieved by using a hybrid student model bringing together features of different
techniques of student modeling.
3.2 Related Work
Each student modeling technique considers, usually, only one or a limited num-
ber of students' characteristics. However, a student model should consider a sig-
nificant number of student's characteristics to be effective. Therefore, the need
Search WWH ::
Custom Search