Student Modeling for Personalized Education: A Review of the Literature - Advances in Personalized Web-Based Education - page 17

(Alfonseca et al. 2006) have modeled the student's learning styles using an overlay
approach. Fuzzy techniques have been, also, used by Stathacopoulou et al. (2005)
for modeling the student's learning style. They have applied a student model to a
discovery-learning environment that aimed to help students to construct the con-
cepts of vectors in physics and mathematics, which drive pedagogical decisions
depending on the student learning style. Furthermore, Crockett et al. (2013) have
tried to predict learning styles in a conversational intelligent tutoring system using
fuzzy logic. Similarly, Oscal CITS adapts to the student's learning styles incorpo-
rated a fuzzy mechanism (Latham et al. 2014). Also, TADV (Kosba et al. 2003,
2005) includes a student model, which combines overlay with fuzzy logic, to
represent communication styles of individual students, except of their knowledge.
Moreover, in GIAS (Castillo et al. 2009) the appropriate selection of the
course's topics and learning resources are based not only on the student's goals
and knowledge level but also on the student's learning style that is modeled
using stereotypes and machine learning techniques. In addition, many research-
ers, like Bunt and Conati (2003), Parvez and Blank (2008), Schiaffino et al.
(2008), and Bachari et al. (2011) have been used Bayesian Networks to detect a
student's learning style and/or preferences automatically. To perform the same
goal, Hernández et al. (2010) have combined Bayesian Networks with cognitive
theories. In addition, Lo et al. (2012) as well as Zatarain-Cabada et al. (2010)
have used artificial neural networks (learning machine) to identify the student's
cognitive and learning styles correspondingly. Finally, the student's preferences in
Personal Reader (Dolog et al. 2004) and in the tutoring system of Pramitasari et al.
(2009) have been modeled by using ontologies.
Many attempts to model other cognitive characteristics of students except of
learning styles have, also, made. Conati et al. (2002) have tried to model in Andes
cognitive aspects like long-term knowledge assessment, plan recognition, ability to
solve problems and reading latency using Bayesian Networks. In Web-PTV (Tsiriga
and Virvou 2003a, b), which teaches the domain of the passive voice of the
English language, the carefulness of the student while solving exercises is esti-
mated through a hybrid student model, which combines stereotypes with machine
learning techniques. Furthermore, in F-CBR-DHTS (Tsaganoua et al. 2003) the
diagnosis of students' cognitive profiles of historical text comprehension was done
with fuzzy techniques and a stereotype-like mechanism. In TELEOS the student's
cognitive behavior has been explicitly diagnosed through Bayesian Networks
(Chieu et al. 2010). AUTO-COLLEAGUE (Tourtoglou and Virvou 2008, 2012)
uses stereotypes to model the personality of students. Durrani and Durrani (2010)
have considered the student's cognitive abilities in the adaptive C ++ tutor CLT
using stereotypes, also. Jia et al. (2010) have designed an adaptive learning system,
which is based on fuzzy logic and helps learners to memory the content and improve
their comprehension. Peña and Sossa (2010) have used an ontology-based student
model to represent learners' knowledge, personality, learning preferences and con-
tent, and to deliver the appropriate option of lecture to students.
Furthermore, DEPTHS (Jeremi et al. 2012), which is an intelligent tutoring

system for learning software design patterns, models, except of the student's

ć