A SELF-REGULATED LEARNING APPROACH: A MOBILE CONTEXT-AWARE AND ADAPTIVE LEARNING

SCHEDULE (MCALS) TOOL

A Self-Regulated Learning Approach: A Mobile

Context-aware and Adaptive Learning Schedule
(mCALS) Tool
J.Y-K. Yau and M. Joy
Department of Computer Science, University of Warwick, Coventry, UK
AbstractSelf-regulated students are able to create
and maximize opportunities they have for studying or
learning. We combine this learning approach with our
Mobile Context-aware and Adaptive Learning
Schedule (mCALS) tool which will create and enhance
opportunities for students to study or learn in
different locations. The learning schedule is used for
two purposes, a) to help students organize their work
and facilitate time management, and b) for capturing
the users activities which can be retrieved and
translated as learning contexts later by our tool. These
contexts are then used as a basis for selecting
appropriate learning materials for the students. Using
a learning schedule to capture and retrieve contexts is
a novel approach in the context-awareness mobile
learning field. In this paper, we present the conceptual
model and preliminary architecture of our mCALS
tool, as well as our research questions and
methodology for evaluating it. The learning materials
we intend to use for our tool will be Java for novice
programmers. We decided that this would be
appropriate because large amounts of time and
motivation are necessary to learn an object-oriented
programming language such as Java, and we are
currently seeking ways to facilitate this for novice
programmers.
Index TermsContext-aware, Learning Schedule,
Mobile Learning, Learning Java programming,
Learning Objects, Self-Regulated Learning
I.

INTRODUCTION

A self-regulated student can be characterized by

his/her active participation in learning from the metacognitive, motivational, and behavioral point of view,
and the characteristics of such a student coincide with
those attributes of higher-performance and highercapacity students [1]. Montalvo and Torres [1] noted that
a self-regulated student is able to
a) use cognitive strategies to organize, transform,
elaborate and recover information;
b) direct their mental processes toward the achievement
of personal goals through plan and control;
c) show positive emotions towards tasks and a high
sense of academic self-efficacy, and have the ability

to control these to adapt to the requirements of the

task and of the specific learning situation;
d) plan and control the time and effort on tasks, and
create and structure preferable learning environments
such as identifying a suitable place for study and
obtaining help from teachers and students when they
experience difficulties;
e) use strategies to maintain their concentration, effort
and motivation and avoid external and internal
distractions, whilst performing tasks.
Initial investigation for the requirements of a mobile
learning organizer established that there was a demand by
users for institutional support of mobile learning,
especially for timetabling information and providing
course content [2]. Learning organizers have been used in
other mobile learning systems [2, 6], however this has not
been for the purpose of capturing and retrieving users
contexts at a later stage. Several mobile learning projects
are underway in Europe and one definition of mobile
learning which has become prevalent from these projects
is that mobile learning is not only about learning using
handheld computer devices, but about learning across
contexts [3]. It is especially important to capture these
contexts (in one way or another) as previous assumptions
which apply to stationary learning applications no longer
apply to applications which function on mobile devices
[4].
Prekop and Burnet [5] divide contexts into Internal
and External dimensions.
The Internal dimensions include human factors such
as users (emotional/physical state, personal events,
beliefs, previous experiences), social environment
(work context, business processes, communication),
and activities (goals, tasks).
The External dimensions include the physical
environment (light, sound, movement, touch,
acceleration, temperature, air pressure, proximity to
other objects, time), infrastructure and location, and
technological features (device and product design).
In this paper, our Mobile Context-aware and Adaptive
Learning Schedule (mCALS) is described as a learning
tool which we believe can be effective for self-regulated
learners as the learning schedule can a) help them
organize their work and facilitate time management, and
b) be used for capturing and retrieving contexts and
allowing our tool to create and enhance opportunities for

A SELF-REGULATED LEARNING APPROACH: A MOBILE CONTEXT-AWARE AND ADAPTIVE LEARNING

SCHEDULE (MCALS) TOOL
students (who are willing to learn) to learn in various
locations. Self-regulated learners require both will and
skill for achievement of learning/studying [1]. Our
mCALS tool aims to provide the skill part for students by
determining which learning materials (in terms of
learning objects) would be appropriate for them at that
location with those contextual attributes which can affect
their ability to learn/study, and we believe that by taking
these
attributes
into
consideration,
students
learning/studying process can be enhanced.
Our research for this model is motivated by the
current lack of pedagogic knowledge in how different
contexts in mobile environments and the implications that
they can bring towards students learning/studying; and
whether by bringing in contexts and making the mobile
learning system context-aware can increase the learning
effectiveness of such mobile learning systems.
Our paper is structured as follows -- in section 2, a
literature review is presented, and the resulting
outstanding problems and issues discussed. In section 3,
the conceptual model and the preliminary architecture of
our mCALS tool is illustrated and discussed, and our
research questions and methodology for evaluation are
then examined. Finally in section 4, we present our
conclusions and suggestions for future work.
II.

LITERATURE REVIEW

A number of context-aware mobile learning tools

have been proposed by different authors. A locationaware learning reminder/organizer was developed by Ryu
and Parsons [6] to help students find their way round the
university campus -- for example, to go to a particular
location for a lecture, -- and also contextual help is
provided when they walk past certain buildings such as
the library books of the students which have become
available are notified to the student. A mobile locationaware handheld event planner was designed and
evaluated by Fithian et al. [7], which allows the user to
view which colleagues are located near him/her, if he/she
decided to have a spontaneous meeting. Normal events
can also be scheduled and notifications sent to
participants.
A system prototype has been developed by Bomsdorf
[8] allowing learning materials to be selected depending
on a given situation this takes into account learner
profiles such as their location, time available for learning,
concentration level and frequency of disruptions.
Similarly, a situation-aware framework/mechanism has
been developed by Bouzeghoub et al. [9] which takes
into account time, place, user knowledge, user activity,
user environment and device capacity for adaptation to
user. A Java Learning Object Ontology has been
developed by Lee et al. [10] as an adaptive learning tool
to facilitate different learning strategies/paths for
students, which can be chosen dynamically.
Existing problems/issues relating to contextawareness (whether within the mobile learning paradigm
or not) include the following.

How to capture users contexts without the users

having to provide them and interfering with what they
are doing, so that it becomes implicit [11]?
How to make the mobile devices aware of the
environment that they are situated in, for example, the
level of noise, lighting, temperature, without attaching
a
number
of
sensors
to
the
devices?
Wearable/pervasive computing can provide a solution
for detecting attributes within the users environment
such as temperature, light; however the wearable
computing might cause inconvenience or discomfort
to the users [11].
How to detect peoples emotions and intentions?
Currently this is still a very difficult task and some
sensors might be able to generate these, however
these may not be entirely accurate [12].
How to maintain users privacy and personal integrity
with the use of location-aware technology knowing
their location [13]?
III. OUR MOBILE CONTEXT-AWARE AND
ADAPTIVE LEARNING SCHEDULE (MCALS) TOOL
Figure 1 depicts the conceptual model of our Mobile
Context-aware and Adaptive Learning Schedule
(mCALS) tool, consisting of three components Learners Schedule/Profile, Adaptation Mechanism,
and Learning Object Repository.
The goal of the system is to select appropriate
learning objects for students based on their current user
contexts and user preferences. The user context attributes
include their location (we use the elements concentration
level and frequency of interruption), and the user
preferences attributes include their knowledge level for
that topic in our case Java -- and their available time.
The user contexts are captured via a learner schedule
(otherwise known as a learning organizer), which is a
novel way of capturing contexts in such a context-aware
mobile learning system. We believe that this learning
schedule approach can be a successful time management
technique and an effective self-regulated learning
approach for motivated students.
Figure 1 Conceptual Model of mCALS

User Contexts
- Available
Time, Type of
Location

Learners
Schedule/
Profile

User
Preferences Knowledge
Level

Learning Object
Repository

Adaptation
Mechanism
Appropriate
learning object(s)
based on the Users
Contexts and Users
Preferences

A SELF-REGULATED LEARNING APPROACH: A MOBILE CONTEXT-AWARE AND ADAPTIVE LEARNING

SCHEDULE (MCALS) TOOL
Figure 2 - System Architecture of the Mobile Context-aware and Adaptive Learning Schedule tool
Learner Model Layer
Learner Profile
KnowledgeLevel
UserID
Name

Learner Schedule

Text File

Student
Database

UpdateKnowledgeLevel
StoreCompletedLOs ()
StorePendingLOs ()
DisplayCompletedLOs ()

ICS File

Nature of Event
Location
Time Start
Time Finish

Software Verification

User Verification

Retrieve GPS ()
Retrieve Wireless
LAN ()
Update Schedule ()

CheckLocation ()
CheckAvailableTim
e ()
Update Schedule ()

Adaptation Layer
Retrieve Contextual Info

Adaptation
Mechanism

RetrieveLocationAttributes ()
RetrieveAvailableTime ()
UserConfirmsContextualAttrib
utes ()
UpdateContextualInfo ()

LO
LO

Recommend
LO ()

Learning Object
Repository

Learning Objects Layer

A. Our Preliminary Architecture of mCALS

Our system architecture of our tool, illustrated in
Figure 2, is logically divided into three layers Learner
Model layer, Adaptation layer and Learning Objects
layer. (Note that this is an expanded version originally
illustrated in [14]).

These events a re stored into the learner schedule and for

the purpose of retrieving the event details with ease, the
calendar will be transformed into an ICS text file.
Figure 2.1 - Input for the Learner Profile and
Learner Schedule Components
Knowledge Level: {Novice,
Intermediate, Advanced}

Learner
Profile

Learner Model Layer

There are three system components within this layer
Learner Profile,
Learner Schedule,
UpdateKnowledgeLevel.
Two methods are in place to check that the Learner
Schedule is up-to-date, namely Software Verification and
User Verification. The Student Database stores details
about the learner profile, and their schedule, and the
students updated knowledge level of Java. For the
learner profile, their knowledge level of Java novice,
intermediate or advanced (in the form of a string); this
will be updated when they have successfully attempted
some learning materials. A unique user ID will be created
for each student, and the students personal details,
although not essential, can also be entered here (such as
name, age, gender) if required.
For the learner schedule, students are required to enter
their scheduled events; the primary attributes being the
nature of the event (such as lecture), location, time starts
and ends (in the form of a string). A graphical-based
calendar will be displayed here for ease of entry.

Nature of Event: {lecture, exam, lunch,

meeting}
Location: {campus, home, restaurant..}
Time Start
Time Finish

Learner
Schedule

Before the Learner Schedule information is retrieved

and accessed by the Retrieve Contextual Info component,
two verification methods software and user verification
are in place to check whether the learners schedule is
still accurate.
For the software verification, we expect that GPS will
detect the appropriate outdoor locations and
WirelessLAN will detect the indoor locations. Therefore
two methods RetrieveGPS() and RetrieveWirelessLAN()
will be in place to retrieve this information, feed it to our
schedule and update where necessary. Using this
information, the learners actual location can be identified
and this can be used to confirm whether the learner is
keeping to his/her schedule.
The user verification will prompt the user to check a)
his/her location and b) his/her available time, and this
information will be used to update the schedule, if
necessary.

A SELF-REGULATED LEARNING APPROACH: A MOBILE CONTEXT-AWARE AND ADAPTIVE LEARNING

SCHEDULE (MCALS) TOOL

Adaptation Layer
Pre-adaptation Retrieve Contextual Information
Figure 2.2 RetrieveLocationAttributes and
RetrieveAvailableTime Methods
Location

Time Start
Time Finish

RetrieveLocation
Attributes ()

RetrieveAvailable
Time ()

Concentration
Level
FrequencyOf
Interruption
AvailableTime

We use the location attribute to calculate two default

values for the level of concentration and frequency of
interruption typical for that type of location. The values
of these attributes in relation to the location have been
obtained by a study performed by Cui and Bull [15]
where they found that different students perceived the
same level of concentration and frequency of interruption
in the same location, although the noise levels may have
been different. These will be used as default levels for the
students concentration level and frequency of
interruption which will be used as grounds for adaptation
of the recommended learning objects (amongst other
factors). Using the Time Start and Time Finish attributes,
we obtain the available time that the student has at a
particular point in time. See Figure 2.2.
Figure 2.3 shows the input and output of the Retrieve
Contextual Info component, the function of which is to
first retrieve the contextual information from the learner
schedule, and then transfer this into actual approximate
values which can be used by the Adaptation Mechanism.
The attributes taken from the learner schedule include
Location, Time Start and Time Finish.
Figure 2.3 Input and Output of the Retrieve
Contextual Info Component

Learner
Schedule
Adaptation
Mechanism
Knowledge Level

Learner Profile

Location
Time Start
Time Finish

Retrieve
Contextual
Info

Concentration Level: {low,

medium, high}
Frequency of Interruption: {low,
medium, high}
Time Available

A method UserConfirmsContextualAttributes() will

also be implemented to give the user the option to view
and confirm the values of these attributes, or change these
values, if necessary. These attributes include
Concentration level, Frequency of Interruption and
Available Time. The method UpdateContextualInfo() will
be used to update this contextual information, as
necessary. Figure 2.2 shows the input and output of the
RetrievedLocationAttributes()
and
RetreiveAvailableTime() methods.

The parameters fed into the mechanism include

knowledge level, concentration level, frequency of
interruption, and available time. Specific adaptation rules
are defined within the adaptation mechanism which will
then take the parameters into consideration and select the
appropriate learning objects for the learner. The Learning
Object Repository contains learning objects for Java, in
the form of text for factual information and examples and
multiple choice questions such as exercises and tests.
Graf and Kinshuk [16] noted that active learners prefer
testing and experimenting and the use of exercises and
tests would be ideal for them; whereas for reflective
learners, they prefer to read materials and therefore
content containing objects and examples would be
appropriate for them.

Learning Objects Layer

The learning objects which have been recommended
to the students will be stored along with the following
information

Whether the student has completed it.

In the case of a test or exercise, whether the
student answered it correctly.

This information will be transferred to the student

database and when the student has attempted an
appropriate amount of material (and accurately), their
knowledge level will be increased.
B. Our Research Questions
Our research questions are divided into five sections
Learning Schedule, Learning Environment/Space,
Contextual Attributes, Our Adaptation Mechanism and
Usability of our tool. We defined Learning Schedule as a
set of activities related to both learning and non-learning
events which have been planned by the user to be
performed.
Learning Schedule
We want to examine whether students use a learning
schedule for time management. One group we will target
for our experiment will be first year computer scientists
because we are currently seeking ways to facilitate
novices to learn Java programming. We want to find out
whether students find that using a learning schedule is an
effective way of managing their time.
How likely is it that these students observe/follow
their learning schedule?
If students have not previously used a learning
schedule, do they think that they can benefit from
using such one for managing their time?
If they do not use one, is that because they foresee
that they cannot observe/follow their learning
schedule?
Are
there
any
particular
personal
characteristics/learning styles relating to the student

A SELF-REGULATED LEARNING APPROACH: A MOBILE CONTEXT-AWARE AND ADAPTIVE LEARNING

SCHEDULE (MCALS) TOOL
which affect whether they use a learning schedule for
time management, and their ability to observe and
follow it? (Characteristics such as conscientiousness,
motivation)
Are students willing to provide their daily learningrelated and non-related activities which are accurate
and up-to-date and which they will conform to?
Learning Environment/Space
Which
places
do
first
year
computer
science/programming
students
learn/study/do
programming in?
Are there specific places which they will go to
perform certain learning activities, such as
programming must be in computer lab?
For each of these places, how well can they
learn/study/do programming in?
Are there restrictions in any of these places which
hinder their ability to perform a certain learning task,
or with the mobile devices that they are using?
Which factors in each of these places affect their
concentration level for learning/studying/doing
programming and to what degree?
How well can students learn/study/do programming if
the place that they are in is prone to frequent
interruptions?
How do interruptions affect the students in
performing certain tasks?
Contextual Attributes (Knowledge Level, Location Concentration Level and Frequency of Interruption,
Available Time)
Is it motivating for computer science students to
learn/study if they are using learning materials which
are adapted and are appropriate for their knowledge
level?
Do they think that the location that they are
learning/studying/doing programming in can affect
how well they learn/study/do programming?
What are the factors within the location which can
affect their ability to study (such as noise level,
movement)
and
how
do
these
affect
learning/studying?
If students are learning in a place prone to
interruptions, would it be helpful for them if the
learning materials selected for them had greater
granularity so that they can frequently stop as
necessary?
Adaptation Mechanism
Would students find it more convenient and more
helpful to their learning when learning materials
selected for them are appropriate for them in the given
contextual situation that they are in?
Will they be able to carry out the activities selected
for them?

Do they think it would be more helpful towards their

learning/studying/doing programming if their
concentration level (to approximation) and/or the
frequency of interruption in the location that they are
situated in and/or the available time that they have for
studying during that period of studying time is taken
into consideration?
Are students able to study/learn/do programming
more effectively (or with more ease of use) when
these contextual features are taken into consideration?
Usability of our tool
Would students use such a device for learning in
various places?
Would students find it convenient and time-saving
that the learning materials selected for them are
appropriate for them in the given contextual situation
that they are in?
Do they like having activities selected for them?
Would students prefer to input the location and time
available manually themselves at the point of using it,
rather than keeping a learning schedule for the system
to retrieve the contexts later when necessary?
Would the students mind the tool knowing where they
are?
C. Our Methodology for evaluating mCALS
Our evaluation methods are divided into two phases
an interview research activity and a system prototype
evaluation. We have currently completed phase 1 of our
research activity and are analyzing the interview data. 40
students participated in our semi-structured interview and
each interview took approximately 20 minutes to
complete and four areas were explored a) studying in
different environments, b) personal information
management, c) learning characteristics, and d) learning
preferences. Phase 2 of our research activity will be
underway once we have completed our interview data
analysis and obtained a deep understanding of what
students required i.e. forming the user requirements of
our context-aware mobile learning system tool. A system
prototype will then be implemented and evaluated via the
means of simulation tests for checking the technical
workings of the system including that that adaptation
mechanism is able to accurately retrieve the up-to-date
context and select appropriate materials for learners based
on their current contexts. We will also ask a number of
students to evaluate our tool in order to obtain learnercentered feedback via questionnaires or interviews.
D. Preliminary Interview Data Analysis
Preliminary analysis of our interview data shows that
the participants were able to identify places which were
most effective for them for studying and had their
preferred learning environment(s), whether in the library,

A SELF-REGULATED LEARNING APPROACH: A MOBILE CONTEXT-AWARE AND ADAPTIVE LEARNING

SCHEDULE (MCALS) TOOL
computer lab, or their bedroom etc. There were many
differences between the students preferred learning
environments, for example, one student found that he/she
was able to concentrate very well in the library because
of the quietness, but another found it extremely
distracting with lots of people around. Factors which the
students noted could distract them from concentration
included noise (whether constant or irregular),
temperature, number of people around, and their
proximity to them.
In terms of personal information management, most
of our participants used a diary for managing their studies
and noted that it was effective for time-management
especially since it reminds them not to forget important
meetings and deadlines. It allows them to schedule their
workload, assign time for specific modules to maximize
effectiveness and efficiency of their studies and some of
the participants highlighted the fact that they were still
able to fit in their social and leisure activities. Our
findings show that participants who used a diary regularly
were fairly likely to keep their scheduled events unless
something unexpected or more urgent came up. On the
other hand, it was also noted by a few of our participants
that they used their diaries as a reminder tool to make
them aware or remind them of possible events/tasks that
they can attend/perform if they had sufficient time such
that not everything in the diary would be strictly adhered
to.
IV.

CONCLUSIONS/FUTURE WORK

In this paper, we have described our Mobile Contextaware and Adaptive Learning Schedule tool as a system
which could aid self-regulated learners whilst they are
learning in mobile learning environments. Their learning
opportunities at different locations could be created or
enhanced by our tool. The originality of our research
includes the capturing of different learning contexts
which are created in different environments and the use
of our learning schedule for storing and retrieving this
information. We plan to expand the scope of our research
not limited to self-regulated learners. Thereafter, we hope
to generalize these requirements for a generic contextaware mobile learning system.
ACKNOWLEDGMENT
This journal paper also appears in the proceedings of
the International Conference on Interactive Mobile and
Computer Aided Learning, April 2008.
REFERENCES
[1] Montalvo, F. and Torres, M. (2004) Self-Regulated
Learning: Current and Future Directions. Electronic Journal
Research in Educational Psychology, 2 (1), 1-34.
[2] Corlett, D. Sharples, M., Bull, S. and Chan, T. (2005)
Evaluation of a Mobile Learning Organiser for University

Students. Journal of Computer Assisted Learning, Vol. 21,

No. 3, pp. 162-170(9).
[3] Sharples, M. (2006) Big Issues in Mobile Learning,
Kaleidoscope Network of Excellence Report.
[4] Lavoie, M. (2006) I, Mlearning: Identifying design
recommendations for a context-aware mobile learning
system, IADIS International Conference on Mobile Learning
, pp. 265-269.
[5] Prekop, P. and Burnett, M. (2003) Activities, context
and ubiquitous computing. Computer Communications,
Vol. 26, Num. 11, pp. 1168-1176.
[6] Ryu, H. and Parsons, D. (2008) A learner-centred design
of a location-aware learning reminder, Int. J. Mobile
Learning and Organization, to appear.
[7] Fithian, R., Iachello, G., Moghazy, J., Pousman, Z. and
Stasko, J. (2003) The design and evaluation of a mobile
location-aware handheld event planner, International
Conference on Human Computer Interaction with mobile
devices and services, pp. 145-160.
[8] Bomsdorf, B. (2005) Adaptation of Learning Spaces:
Supporting Ubiquitous Learning in Higher Distance
Education, Dagstuhl Seminar Proceedings 05181 Mobile
Computing and Ambient Intelligence: The Challenge of
Multimedia.
[9] Bouzeghoub, A., Do, K. and Lecocq, C. (2007)
Contextual Adaptation of Learning Resources, IADIS
International Conference Mobile Learning, pp. 41-48.
[10] Lee, M., Ye, D. and Wang, T. (2005) Java Learning
Object Ontology, International Conference on Advanced
Learning Technologies, pp. 538-542.
[11] Schmidt, A. (2002) Ubiquitous Computing
Computing in Context. PhD thesis.
[12] Schmidt, A. (2000) Implicit Human Computer
Interaction Through Context. Personal Technologies, Vol
4(2), June 2000.
[13] Synnes, K., Nord, J. and Parnes, P. (2003) Location
Privacy in the Alipes Platform, International Conference on
System Sciences, pp. 10-19.
[14] Yau, J. and Joy, M. (2007) Architecture of a Contextaware and Adaptive Learning Schedule for Learning Java,
International
Conference
on
Advanced
Learning
Technologies, pp. 252-256.
[15] Cui, Y. and Bull, S. (2005) Context and learner
modeling for the mobile foreign language learner. Science
Direct, System 33, pp. 353-367.
[16] Graf, S. and Kinshuk (2006) An Approach for detecting
learning styles in learning management systems,
International
Conference
on
Advanced
Learning
Technologies, pp. 161-163.

AUTHORS
J. Y-K Yau is with the University of Warwick (e-mail:
j.y-k.yau@ warwick.ac.uk).
M. Joy is with the University of Warwick (e-mail:
[email protected]).
Manuscript received 15 November 2007.