2014 International Conference on Information, Communication Technology and System
2014 International Conference on Information, Communication Technology and System
Towards an Adaptive Model to Personalise Open
Learning Environments using Learning Styles
Heba Fasihuddin, Geoff Skinner and Rukshan Athauda
The University of Newcastle
Faculty of Science and Information Technology
Callaghan, Australia
Abstract— Open learning represents a new form of online
learning. It is based on providing courses, learning materials for
free to be taken by any interested learner. The current model of
open learning has certain limitations which provide potential for
improvement. One such area is personalization in learning
environments. One avenue to enhance learning experience in
open learning environments is giving consideration to learning
principles and cognitive science. This paper aims to introduce a
proposal for an adaptive model to personalize the open learning
environments based on the theory of learning styles and
particularly the Felder and Silverman Learning Style Model
(FSLSM). This model consists of two main agents to perform its
functionalities. First, the identification agent which is responsible
of identifying the learners’ learning styles by monitoring certain
determined patterns of learners’ behaviors with learning objects
while the learner interact with learning materials. Second, the
recommender agent which is responsible of providing an
adaptable navigational support based on the identified learning
styles and preferences. The paper presents a description of the
model and its functionalities including the patterns that can be
monitored in open learning environments to identify the learning
styles and also how the adaptation support can be provided based
on the identified styles. Future implementation will test and
verify this proposed model.
Keywords— Adaptive systems; learning styles; MOOCs; open
learning; personalization.
I. INTRODUCTION
Online learning is an evolutionary learning approach that
keeps evolving and changing due to the continuous evolution
of technology. Open learning is a new phenomenon of online
learning that allows learning materials to be freely available on
the Internet for anyone who is interested. This new
phenomenon becomes a tangible reality due to the newly
emerged cloud computing technology.
Recently, various prestigious learning institutions, such as
Harvard, MIT, and Stanford, provide learning materials in an
open approach. Coursera [1], edX [2], Udacity [3], Udemy [4],
and many others are examples of this inventive open learning
style. Courses that are provided through these open learning
environments are termed Massive Open Online Courses
(MOOCs). Despite the popularity of MOOCs and the fact that
they attract an enormous number of learners [5], there are
certain limitations that still need to be considered and managed
978-1-4799-6858-9/14/$31.00 © 2014 IEEE
Authorized licensed use limited to: University of Peshawar. Downloaded on September 17,2020 at 08:51:39 UTC from IEEE Xplore. Restrictions apply.
to enhance such an open model of learning. These limitations
relate to different aspects of the courses, such as teaching and
learning methods; learning content; assessments; identity
authentication; accreditation; and learners’ varying needs,
among others. All of these limitations raise different concerns
about the sustainability of open learning. The authors believe
that there is a need to enhance the current model of open
learning, and find one of the avenues is by considering
cognitive science and learning principles [6]. The authors have
previously selected certain learning theories that can be
considered to enhance the presentation and organization of
learning materials in open learning environments, and also to
personalize the learning experience and adapt to individual
learners’ needs and preferences [7].
The focus of this paper is the personalization of open
learning environments based on the theory of learning styles.
Learning style refers to the way a learner receives and
processes information; therefore, every learner has a different
learning style [8]. Among different existing models of learning
styles, the Felder and Silverman Learning Style Model
(FSLSM) has been selected to be incorporated in open learning
environments to provide a personalized learning experience.
The FSLSM has been considered in different studies to
provide adaptive learning environments. Some of these studies
were based on using the Index of Learning Styles (ILS) [9],
which is a questionnaire developed by Felder and Soloman to
identify the learning style in order to provide learning objects
that suit each learner’s learning style. Other studies were
concerned with introducing a mechanism to automatically
identify the learning styles [10, 18, 24, 25, 31]. Different
approaches were applied and different mechanisms have been
introduced in the literature. More details about these
approaches will be provided later in this paper.
This paper aims to introduce a proposed adaptive model to
personalize the open learning environments based on learners’
preferred learning styles. The goal of this model is to identify
the learners’ learning styles and consequently provide a
personalized navigational support. To automatically identify
learning styles, it has been proposed to apply the literature-
based approach [10]. This is because this approach has been
shown to have higher accuracy results in detecting learning
styles [10]. It is mainly based on monitoring the learners’
behaviors on some determined patterns based on the FSLSM.
ICTS 2014, Surabaya, Indonesia 183
These patterns are determined based on the learning objects
that are common in some open environments, such as edX,
Coursera, Udemy, and Udacity. The proposed model considers
the learning objects in open learning environments which is a
distinction from other studies that consider blended learning
environments.
The rest of this paper is organized as follows: first, a
background of the related concepts is presented in section 2.
Section 3 provides an overview of adaptive systems with
respect to learning styles. Section 4 presents a review of
previous work on adaptive systems based on learning styles.
The incorporation of the FSLSM to open learning
environments is presented in section 5. Finally, the paper is
concluded in section 6 with a brief overview of future work.
II. BACKGROUND
A. Open Learning Environments
As mentioned, the evolution of technology leads to
continual change and development in online learning
approaches, and recently, open learning has emerged as a new
form of online learning based on the utilization of cloud
computing capabilities. In open learning, courses are freely
available on the Internet to be accessed by anyone who is
interested. These courses are provided by different learning
providers who could be academics representing learning
institutions or individuals who have appropriate knowledge and
expertise. Recently many courses have been offered in this
form by different prestigious institutions, such as Stanford and
MIT. These courses have gradually refined into what are
known as Massive Open Online Courses (MOOCs).
MOOCs offer free university-level courses online and have
two key features – open access and scalability [11]. These two
features allow MOOCs to be taken online by anyone and
enable the courses to be designed to support an indefinite or
even infinite number of participants. They are learner-centered
courses, so learners are able to work and learn at their own
pace, which gives them the opportunity to maintain their peak
cognitive level and ability as they are able to access learning
resources repeatedly until they meet their learning needs.
MOOCs are a current point of controversy, and their
success cannot yet be determined. There are some limitations
of MOOCs that make them a point of contention. These
limitations relate to various factors including accreditation,
authentication, teaching methods, and subjects that can be
taught in such a way [12]. Moreover, it has been claimed that
MOOCs place less emphasis on providing interactive and
dynamic approaches to learning, as there is no creativity in
delivering learning content, only the use of traditional
approaches (i.e. video lectures) that lack support for learners’
variable needs [13].
Based on the highlighted dilemmas, the authors believe that
there is no clear frontrunner in terms of a model for MOOCs or
open learning environments. Rather, MOOCs and other open
learning initiatives are in their early stages of evolution.
Different aspects and research questions still need to be
considered and addressed in order to achieve a valid model [7].
The primary author and related research is sponsored by King Abdulaziz
University in Saudi Arabia.
184
Authorized licensed use limited to: University of Peshawar. Downloaded on September 17,2020 at 08:51:39 UTC from IEEE Xplore. Restrictions apply.
However, there is the possibility that a single model might not
be a solution to all learning needs and environments, and a
plethora of models and systems may emerge catering to
differing learning needs and environments.
The development of open learning environments is a
critical and sensitive field due to the implications they have on
learners, instructors, and the learning process. Therefore,
scientific principles for learning should be considered in the
development of MOOCs in order to achieve the desired
learning goals. It is stated in [6] that tailoring general learning
principles and working with cognitive scientists is one
approach that needs to be considered to enhance MOOCs and
provide the best outcomes for learners. Based on this, the
authors consider the theory of learning styles [8] to introduce
an approach for personalizing open learning environments.
This is believed to increase learners’ satisfaction and lead to
better learning outcomes. Following is an overview of this
theory and its implications.
B. The Theory of Learning Styles
Learning style refers to the way a learner receives and
processes information [8]. Therefore, different learners have
different learning styles [8]. Considering learning styles in the
courseware design has been found to be effective and
beneficial in learning. It has been stated that providing learners
with learning materials and activities that suit their preferences
and learning styles makes learning easier for them [14]. This
statement has been proven by many studies that found that
students can achieve better learning outcomes and higher
scores [15], and can also master the learning materials in less
time [16].
In literature, several models for learning styles were
defined and found to be valid and reliable [17]. However, the
Felder and Silverman Learning Style Model (FSLSM) is
selected to be applied to personalize open learning
environments [8]. This selection is based on recommendations
in the literature [18, 19, 20]. The FSLSM was mainly proposed
for engineering education, and according to Felder, the original
paper of their proposed model was the most frequently cited
paper in articles published in the Journal of Engineering
Education over a 10-year period. In addition, the mechanism of
its Index of Learning Style (ILS) questionnaire that was
developed by Felder and Soloman to identify learning styles
can be easily applied to adaptive systems [9]. Furthermore, it
has been stated that the FSLSM is the most appropriate and
feasible model to be implemented in adaptive courseware [18,
19]. Moreover, a study that was conducted to compare the
suitability of different learning style models to be applied in
online learning concluded that the FSLSM is the most
appropriate model [20]. This is because it provides more details
that enable it to be tailored to individual needs and preferences,
providing adaptability. Considering this, the authors
hypothesize that such a model will be effective in online
learning and particularly in open learning.
The FSLSM classifies learning styles into four dimensions
and identifies two types of learners for each dimension. The
dimensions are perception, input, processing, and
understanding. Firstly, the perception dimension defines the
type of information that learners prefer to receive and learn by;
intuitive learners prefer meaning and theories, while sensory
learners prefer learning by examples and practice. The second
dimension is input, which defines the approach the learners
prefer to learn with; visual learners prefer pictures, diagrams,
and flowcharts, while verbal learners prefer written or spoken
explanations. The processing dimension indicates how learners
prefer to process and practice their learning; active learners
prefer working with others, while reflective learners prefer
thinking and working alone. Finally, the understanding
dimension indicates how learners progress toward
understanding; sequential learners learn in continual small
steps, while global learners learn holistically in large jumps.
Table I represents these styles and their associated types.
TABLE I. FELDER AND SILVERMAN LEARNING STYLES
Dimension Preferred Learning Styles
Perception Sensory Intuitive
Input Visual Verbal
Processing Active Reflective
Understanding Sequential Global
III. ADAPTIVE SYSTEMS WITH RESPECT TO LEARNING
STYLES
Adaptive systems have been described as the systems that
are able to provide personalized learning support to the learner
throughout their interaction based on the goals, preferences,
and knowledge of each individual learner [21]. It has been
found that adaptive learning systems lead to better learning
outcomes, reduce time and effort required, and increase
learners’ satisfaction [22]. Adaptive systems can adapt to user
data, usage data, and environments data [21]. User data refers
to various characteristics of the user, such as learning styles
and cognitive traits. Usage data refers to user interaction with
the systems. Environment data refers to the adaptation to user
context, including location or platform. Providing adaptability
based on the considered factors has been classified into two
different areas – adaptive presentation and adaptive navigation
support [21]. Adaptive presentation comprises text and
multimedia adaptation technologies, while adaptive navigation
support comprises links sorting, hiding, and direct guidance.
Systems that are adaptive to learning styles need to identify
the learner’s learning style first and then adapt to the learner’s
preferences. Adaptation methods of adaptive systems have
been classified into two different approaches – collaborative
and automatic [23]. In the collaborative approach, learners are
asked to provide their preferences explicitly by taking a test or
filling out a questionnaire, such as the ILS questionnaire [9], in
order to build their adaptable models. On the other hand, in the
automatic approach, the learners’ adaptable models are built
automatically by the adaptive systems through intelligent and
machine learning techniques that exploit learners’ interactions
and behaviors while they are using the systems for learning.
In the literature, two different methods for identifying
learning styles based on the FSLSM were used – the data-
driven method and the literature-based method [10]. Both
methods rely on some identified patterns to detect the learning
style of the learner. These patterns are based on monitoring the
provided learning objects in such a way that they adhere to the
Authorized licensed use limited to: University of Peshawar. Downloaded on September 17,2020 at 08:51:39 UTC from IEEE Xplore. Restrictions apply.
FSLSM. The data-driven method aims to build a model that
imitates the ILS questionnaire and uses sample data to
construct a model. Some of the techniques used to apply this
method are neural network, decision tree, Hidden Markov
Model, fuzzy clustering, and Bayesian network. The literature-
based method uses the behavior of students and actions with
the systems while they are learning in order to identify their
learning style preferences. It is entirely literature-based, as
patterns are identified based on findings of learners’
preferences and behaviors for each specific learning style. This
method uses only simple rule-based methods to calculate
learning styles. A study has been conducted to compare the
efficiency of these two methods in detecting learning styles,
and has found that the literature-based method gives more
accurate results than the data-driven method [10]. Although the
literature-based method has been found to be efficient, it has
been claimed in [24] that this method’s point of weakness is
embodied in the possibility of not considering all the potential
patterns that could affect the detection of learning styles. Many
studies have been conducted to automatically identify learning
styles and following is an overview of them.
IV. RELATED WORK
Building adaptive systems that adapt to learners’ learning
styles has been a point of interest in research. Different studies
have been done to provide adaptive learning based on learning
styles. Some of these studies were based on the collaborative
adaptive approach, where students were asked to provide their
preferences through the answers to the ILS questionnaire;
while others were based on the automatic approach, where their
learning styles were detected automatically through their
behaviors and interactions with the systems. In the literature, a
variety of methods and techniques were used. These methods
differ based on the attributes that were used for detecting
learning styles (personality factors, behavior factors), the
underlying technique (literature-based, data-driven) and the
underlying infrastructure (Learning Management Systems
(LMSs), special user interface).
Various studies have considered the variety of learning
styles and the importance of incorporating them into learning
environments. Some of these studies were concerned with
introducing models and approaches to incorporate the FSLSM
into the adaptive systems based on the collaborative approach
[19, 25, 26]. Recently, studies have been more concerned with
automatically detecting the learners’ learning styles rather than
using the collaborative approach. As stated previously in this
paper, two main approaches are found in the literature – the
data-driven approach and the literature-based approach. These
two approaches were implemented using different metrics and
techniques. In addition, different patterns and learning objects
were considered in a way that suits the adopted learning
environments. In the data-driven approach, some data mining
and machine learning algorithms were used to automatically
identify the learners’ learning styles. Some examples are
Bayesian network [27, 28], neural network [29, 30], decision
tree [31], NBTree [32], Hidden Markov Model [31], k-nearest
neighbor algorithm along with genetic algorithm [33], and
AprioriAll mining algorithm [34]. Graf was the first to use the
literature-based approach to automatically identify learning
185
styles [10, 35]. She determined different patterns of learners’
behaviors and actions based on the common learning objects in
LMSs to identify learning styles using the simple rule-based
method. Other studies have also used this approach to identify
some dimensions of the FSLSM [24, 36, 37]. Some other
recent studies have considered metrics other than monitoring
the behaviors on learning objects. For example, the study
presented in [38] found that the mouse vertical speed can be
used as an indication of global/sequential learning styles.
Another example is the use of biometrics technologies, such as
eye trackers, to identify visual/verbal learning styles [39].
However, these biometrics cannot be used in open learning
environments as they require special equipment and devices
that might not be available to all learners.
V. INCORPORATING THE FSLSM INTO OPEN LEARNING
ENVIRONMENTS
Currently, open learning environments are based on the
“one size fits all” approach. There is no personal support or
consideration of individual needs and preferences. As open
learning environments target a massive number of learners with
different backgrounds, preferences, and cognitive abilities,
adaptability to the variability of learners’ needs and preferences
is crucial.
Open learning environments can be personalized and
provide adaptable support by incorporating FSLSM. The
authors’ focus in this paper is to present a model to personalize
open learning environments based on the adaptability to
learners’ learning styles. This model aims to provide adaptive
navigational support to individual learners based on their
preferred learning style. Therefore, this model has two main
functions, which are identifying the learning styles and
providing navigational support. These functions are performed
by an adaptive engine which consists of two agents: 1)
Learning style identification agent 2) Recommender agent. The
design of the identification agent has been based on the
literature-based method, which requires some determined
patterns of learners’ interactions with the provided learning
objects to be monitored in order to identify the learning styles.
The identified styles will be stored in the learners’ profiles in
order to feed the recommender agent and provide the desired
adaptability for learners. In addition, to ensure a dynamic
adaptability, the stored styles will be updated regularly while
learners are learning based on their interactions and behaviors
with the learning objects. An illustration of the proposed
adaptive model is provided in “Fig. 1,”.
A. Patterns to Identify Learning Styles
As mentioned earlier, a set of interaction patterns should be
determined to allow the learning style identification agent to do
its functionality. Determining patterns for identifying learning
styles in open learning environments should be based on the
learning objects in these environments. For that, the authors
have observed the provided learning objects in some of the
well-known MOOCs, such as edX [2], Coursera [1], Udemy
[4] and Udacity [3]. The identified learning objects include
course overviews, outlines, video lectures, number of learning
objects that vary between textual-based and visual-based,
discussion forums, examples, exercises, quizzes with
186
Authorized licensed use limited to: University of Peshawar. Downloaded on September 17,2020 at 08:51:39 UTC from IEEE Xplore. Restrictions apply.
Fig. 1 An illustration of the adaptive model
immediate feedback and additional reading materials.
Felder and Silverman have presented a good explanation
for the FSLSM in terms of learners’ preferences and the
corresponding learning and teaching styles [8]. Considering
this and other literature [24, 31, 35, 37, 40], the authors have
determined patterns to identify learning styles in open learning
environments; these patterns consider the previously listed
learning objects. In addition, knowledge maps have been
considered as a learning object, and this is based on the
suggestion of organizing learning concepts in a knowledge map
as an approach to enhancing the learning experience in open
learning environments [7][41].
To identify the preferred learning styles for each
dimension, certain patterns of behaviors need to be monitored
in relation to pre-determined threshold values. Following are
descriptions of these patterns of behaviors. First, in terms of
learners’ perceptions, sensing learners prefer facts, data and
experimentation (i.e. concrete materials), while intuitive
learners prefer principles and theories (i.e. abstract materials),
so annotating the learning objects to specify their types (i.e.
concrete and abstract) and the learners’ access to these objects
and the time spent on them can be used as a pattern. In
addition, sensing learners like to solve problems by standard
methods and do not like surprises, while intuitive learners like
to invent new ways to solve problems. Based on this, sensing
learners are expected to access more examples and spend more
time on them, while intuitive learners spend more time on the
learning materials. These can be considered other patterns to
distinguish between sensors and intuitive learners. Sensing
learners are patient with details, careful but slow, while
intuitive learners tend to be quick and careless; therefore,
sensing learners spend more time on quizzes while intuitive
learners spend less time.
In regard to the Input dimension, visual learners remember
what they see better than what they listen to, while verbal
learners remember more of what they hear than what they see.
Visual learners learn better by diagrams, flowcharts and
pictures, while verbal learners prefer verbal explanation rather
than visual demonstration. Therefore, annotating the learning
objects to distinguish whether they are visual or verbal, and the
access and time spent on them, can be considered a pattern.
In regard to the Processing dimension, active learners like
to try out and learn by practice, while reflective learners prefer
to think and reflect about what they learn, so they learn better
by observation. Based on this, active learners tend to access
more exercises and spend more time on them. In addition,
active learners like to work in groups, while reflective learners from instructors to provide learning materials in different forms
prefer to learn alone; therefore, active learners access the and also determine the expected time for each, we expect this
discussion forums and posts more than the reflective learners. effort to produce rich open learning environments with the
ability to personalize based on learning styles and preferences.
Finally, in regard to the Understanding dimension, Consequently, this may increase the learners’ satisfaction and
sequential learners like to learn in a sequential process and also their learning outcomes.
prefer learning materials to be organized and presented in a
steady progression of complexity and difficulty. Global
learners do not like the linear approach and might jump directly B. Providing Adaptive Support based on the Identified
to the more complex materials. Based on this, the behavior of Learning Styles
accessing the learning materials can be considered as a pattern. After identifying the learning styles the recommender agent
In addition, global learners like to be provided with the overall will provide an adaptable navigational support. This means that
picture of the provided topic; therefore, they access and spend every learner will get the learning objects organized in such a
more time on the overview and outline. Moreover, global way that suits their learning style. This organization will be
learners are expected to access the knowledge maps of the based on the recommendations in [8, 16]. As mentioned,
learning concepts more than the sequential learners, so the time sensing learners prefer to learn from concrete materials, so
spent on accessing the knowledge maps is another pattern. these types of learning objects need to be shown before the
Table II summarizes all the mentioned patterns to identify abstract materials. The opposite needs to be done for intuitive
learning styles in open learning environments. learners – abstract materials need to be shown to them first. In
addition, sensing learners prefer to learn by examples and real-
life applications, so examples need to be shown to them before
TABLE II. PATTERNS FOR IDENTIFYING LEARNING STYLES IN OPEN
LEARNING ENVIRONMENTS the explanation, while intuitive learners prefer the reverse.
Moreover, sensing learners prefer to get more examples and
Sensing learners Intuitive learners exercises, so all the available examples and exercises need to
• Example access (+) a • Example access (-) b
• Example time (+) • Example time (-)
be recommended to them, while just some can be
• Exercise access (+) • Exercise access (-) recommended to intuitive learners. In terms of the input
• Exercise time (+) • Exercise time (-) dimensions, textual-based learning objects can be
• Details and concrete learning • Details and concrete learning recommended to verbal-based learners, while the visual-based
objects time (+) objects time (-)
• Abstract learning objects time (-) • Abstract learning objects time(+)
objects can be recommended to the visual learners. For the
• Additional reading access (+) • Additional reading access (-) processing dimension, active learners prefer to learn by doing,
• Quiz time (+) • Quiz time (-) so more exercises will be provided to them in between the
Visual learners Verbal learners learning material. They also like to invent their own
• Visual learning objects access(+) • Visual learning objects access (-) approaches to solving problems, therefore, fewer examples will
• Visual learning objects time (+) • Visual learning objects time (-)
Active learners Reflective learners
be shown to them. The reverse approach needs to be taken for
• Exercise access (+) • Exercise access (-) reflective learners, so more examples will be shown in between
• Exercise time (+) • Exercise time (-) the learning objects and less exercises. In regard to the
• Example access (-) • Example access (+) understanding dimension, sequential learners prefer to learn by
• Example time (-) • Example time (+) a linear approach, so learning objects involving examples and
• Forum access (+) • Forum access (-)
• Forum Post (+) • Forum Post (-) exercises need to be organized in a linear increase of
Sequential learner Global learners complexity, and the course conclusion and knowledge map are
• Outline access (-) • Outline access (+) to be shown last. In contrast, the conclusion and the knowledge
• Outline time (-) • Outline time (+) map need to be presented first to global learners.
• Access topic knowledge map (-) • Access topic knowledge map (+)
• Linear access for learning • Linear access for learning Providing learning objects in the previous described
concepts (+) concepts (-) organization is believed to enhance the learning experience in
a.
(+) Indicates high number of occurrence or more time open learning environments and consequently enhance the
b.
(-) Indicates low number of occurrence or less time learners’ satisfaction and learning outcomes.
Monitoring the previously listed patterns of learners’
behaviors requires predetermined threshold values for each VI. CONCLUSION AND FUTURE WORK
pattern in order to evaluate the level of each pattern’s
This paper introduces an idea towards developing an
appearance in learners’ behaviors while learning. We intend to
adaptive model to personalize open learning environments
use various statistical techniques and other means to accurately
based on the theory of learning styles and particularly the
identify thresholds and thereby determine learning styles and
Felder and Silverman Learning Style Model (FSLSM). The
preferences of learners. By comparing ILS survey results with
paper presents a description of the adaptive model components
identified learning styles, we are able to find the accuracy of
along with the determined patterns of learners’ behaviors that
our methodology while also benchmarking across other
can be monitored to identify learning styles. These patterns are
approaches. Our approach has two main overheads. Firstly, the
determined based on the provided learning objects in open
instructors need to provide learning materials catering to
learning environments. In addition, a description of how to
different learning styles (e.g. Visual vs. Verbal). Next, the
provide adaptable navigational support is discussed.
instructors need to identify the expected time for every
provided learning material. Although this model requires effort

187

Authorized licensed use limited to: University of Peshawar. Downloaded on September 17,2020 at 08:51:39 UTC from IEEE Xplore. Restrictions apply.
 Future work of this study involves developing a prototype
that simulates open learning environments in terms of offering
open online courses that learners can take and learn at their
own pace. This prototype will be dynamically adaptable to
learners’ learning styles based on the presented patterns, and
will also provide individual navigational support. Moreover,
evaluations of the precision of identifying learning styles and
the learners’ satisfaction about the provided adaptability will be
conducted.
REFERENCES
[1] Coursera. "Coursera," https://www.coursera.org/. 2012.
[2] edX. "edX," http://www.edxonline.org/. 2012.
[3] Udacity. "Meet Udacity!," http://www.udacity.com/. 2012.
[4] Udemy. "Udemy," https://www.udemy.com/. 2014.
[5] L. Breslow, D. Pritchard, J. DeBoer, G. Stump, A. Ho, and D. Seaton,
"Studying learning in the worldwide classroom: Research into edX's first
MOOC," Research & Practice in Assessment, vol. 8, pp. 13-25, 2013.
[6] J. J. Williams, "Improving learning in MOOCs with Cognitive Science,"
in AIED 2013 Workshops Proceedings Volume, 2013, p. 49.
[7] H. A. Fasihuddin, G. D. Skinner, and R. I. Athauda, "Boosting the
Opportunities of Open Learning (MOOCs) through Learning Theories,"
Journal on Computing, vol. 3, pp. 112-117, 2013.
[8] R. Felder & L. Silverman, "Learning and teaching styles in engineering
education," Engineering education, vol. 78, pp. 674-681, 1988.
[9] B. A. Soloman and R. M. Felder. "Index of Learning Styles
Questionnaire," http://www.engr.ncsu.edu/learningstyles/ilsweb.html.
[10] S. Graf, "Adaptivity in Learning Management Systems Focussing on
Learning Styles," Ph.D. Thesis, Vienna University of Technology,
Austria, 2007.
[11] L. Yuan and S. Powell, "MOOCs and Open Education: Implications for
Higher Education," CETIS JISC, vol. 21, p. 2013, 2013.
[12] J. J. Roberts and J. O'Loughlin. "The Season of the MOOC,"
UNIversitas 8. http://www.uni.edu/universitas. 2013.
[13] C. Parr. "Mooc Creators Criticise Courses’ Lack of Creativity,"
http://www.timeshighereducation.co.uk/news/mooc-creators-criticise-
courses-lack-of-creativity/2008180.fullarticle. 2013.
[14] S. Graf and L. Tzu-Chien, "Supporting Teachers in Identifying Students'
Learning Styles in Learning Management Systems: An Automatic
Student Modelling Approach," Journal of Educational Technology &
Society, vol. 12, pp. 3-14, 2009.
[15] N. Bajraktarevic, W. Hall, and P. Fullick, "Incorporating learning styles
in hypermedia environment: Empirical evaluation."
[16] S. Graf and K. Kinshuk, "Providing Adaptive Courses in Learning
Management Systems with Respect to Learning Styles," presented at the
World Conference on E-Learning in Corporate, Government,
Healthcare, and Higher Education 2007, Quebec City, Canada, 2007.
[17] F. Coffield, D. Moseley, E. Hall, and K. Ecclestone, "Should we be
using learning styles?: What research has to say to practice," 2004.
[18] P. García, S. Schiaffino, and A. Amandi, "An enhanced Bayesian model
to detect students’ learning styles in Web-based courses," Journal of
Computer Assisted Learning, vol. 24, pp. 305-315, 2008.
[19] C. Carver, R. Howard, and W. Lane, "Enhancing student learning
through hypermedia courseware and incorporation of student learning
styles," IEEE Transactions on Education, vol. 42, pp. 33-38, 1999.
[20] J. Kuljis and F. Liu, "A Comparison of Learning Style Theories on the
Suitability for elearning," Web Technologies, Applications, and
Services, vol. 2005, pp. 191-197, 2005.
[21] P. Brusilovsky, "Adaptive hypermedia," User modeling and user-
adapted interaction, vol. 11, pp. 87-110, 2001.
[22] S. Graf and Kinshuk, "Adaptive Technologies," in Handbook of
Research on Educational Communications and Technology, ed: Springer
New York, 2014, pp. 771-779.
188
Authorized licensed use limited to: University of Peshawar. Downloaded on September 17,2020 at 08:51:39 UTC from IEEE Xplore. Restrictions apply.
[23] P. Brusilovsky, "Methods and techniques of adaptive hypermedia," User
modeling and user-adapted interaction, vol. 6, pp. 87-129, 1996.
[24] N. Ahmad, Z. Tasir, J. Kasim, and H. Sahat, "Automatic Detection of
Learning Styles in Learning Management Systems by Using Literature-
based Method," Procedia-Social and Behavioral Sciences, vol. 103, pp.
181-189, 2013.
[25] H. Hong and D. Kinshuk, "Adaptation to student learning styles in web
based educational systems," in World Conference on Educational
Multimedia, Hypermedia and Telecommunications, 2004, pp. 491-496.
[26] J. E. Gilbert and C. Y. Han, "Adapting instruction in search of ‘a
significant difference’," Journal of Network and Computer Applications,
vol. 22, pp. 149-160, 1999.
[27] C. Carmona, G. Castillo, and E. Millan, "Designing a Dynamic Bayesian
Network for Modeling Students' Learning Styles," in 8th IEEE ICALT
'08, 2008, pp. 346-350.
[28] P. García, A. Amandi, S. Schiaffino, and M. Campo, "Evaluating
Bayesian networks’ precision for detecting students’ learning styles,"
Computers & Education, vol. 49, pp. 794-808, 2007.
[29] R. Z. Cabada, M. L. B. Estrada, R. Z. Cabada, and C. A. R. Garcia, "A
Fuzzy-Neural Network for Classifying Learning Styles in a Web 2.0 and
Mobile Learning Environment," in Web Congress, 2009. LA-WEB '09.
Latin American, 2009, pp. 177-182.
[30] A. Latham, K. Crockett, and D. Mclean, "Profiling Student Learning
Styles with Multilayer Perceptron Neural Networks," in IEEE SMC,
2013, pp. 2510-2515.
[31] H. Cha, Y. Kim, S. Park, T. Yoon, Y. Jung, and J.-H. Lee, "Learning
Styles Diagnosis Based on User Interface Behaviors for the
Customization of Learning Interfaces in an Intelligent Tutoring System,"
in Intelligent Tutoring Systems. vol. 4053, M. Ikeda, K. Ashley, and T.-
W. Chan, Eds., ed: Springer Berlin Heidelberg, 2006, pp. 513-524.
[32] E. Özpolat & G. Akar, "Automatic detection of learning styles for an e-
learning system,"Computers & Education,vol. 53, pp. 355-367, 9// 2009.
[33] Y.-C. Chang, W.-Y. Kao, C.-P. Chu, and C.-H. Chiu, "A learning style
classification mechanism for e-learning," Computers & Education, vol.
53, pp. 273-285, 9// 2009.
[34] A. Klašnja-Miliüeviü, B. Vesin, M. Ivanoviü, and Z. Budimac, "E-
Learning personalization based on hybrid recommendation strategy and
learning style identification," Computers & Education, vol. 56, pp. 885-
899, 4// 2011.
[35] S. Graf, Kinshuk, and L. Tzu-Chien, "Identifying Learning Styles in
Learning Management Systems by Using Indications from Students'
Behaviour," in 8th IEEE ICALT '08., 2008, pp. 482-486.
[36] Ö. ùimúek, N. Atman, M. ønceo÷lu, and Y. Arikan, "Diagnosis of
Learning Styles Based on Active/Reflective Dimension of Felder and
Silverman’s Learning Style Model in a Learning Management System,"
in Computational Science and Its Applications – ICCSA 2010. vol.
6017, D. Taniar, O. Gervasi, B. Murgante, E. Pardede, and B. Apduhan,
Eds., ed: Springer Berlin Heidelberg, 2010, pp. 544-555.
[37] N. Atman, M. Inceo÷lu, and B. Aslan, "Learning styles diagnosis based
on learner behaviors in web based learning," in Computational Science
and Its Applications – ICCSA 2009, ed: Springer, 2009, pp. 900-909.
[38] D. Spada, M. Sánchez-Montañés, P. Paredes, and R. Carro, "Towards
Inferring Sequential-Global Dimension of Learning Styles from Mouse
Movement Patterns," in Adaptive Hypermedia and Adaptive Web-Based
Systems. vol. 5149, W. Nejdl, J. Kay, P. Pu, and E. Herder, Eds., ed:
Springer Berlin Heidelberg, 2008, pp. 337-340.
[39] T. J. Mehigan, M. Barry, A. Kehoe, and I. Pitt, "Using eye tracking
technology to identify visual and verbal learners," in IEEE IICME,
2011, pp. 1-6.
[40] S. Graf and S. Viola, "Automatic student modelling for detecting
learning style preferences in learning management systems," in Proc.
CELDA 2009, pp. 172-179.
[41] H. Fasihuddin, G.Skinner and R. Athauda, "Insights into the use of
Knowledge Maps in Online Learning Environments: A Pilot Study," in
Proc. The 1st International Conference on Technical Education
(ICTechEd2013), 2013, pp. 27-32.