CAMELL
Towards a Ubiquitous Multilingual e-Learning System
Maria Virvou and Christos Troussas
Department of Informatics, University of Piraeus, 80, Karaoli & Dimitriou st., 18534, Piraeus, Greece
Keywords: Computer-assisted multiple language learning, Ubiquitous e-learning, Intelligent tutoring systems, User
modelling, Error diagnosis.
Abstract: This paper describes a ubiquitous e-learning tutoring system for multiple language learning. It is a post-
desktop model of human-computer interaction in which students “naturally” interact with the system in
order to get used to electronically supported computer-based learning. The system presents advances in user
modeling and user interface design. Furthermore, the main focus of the tutor is on the student’s error
diagnosis process, which is performed by the student modeling component. Whenever a student types a
solution for an exercise, the system’s reasoning mechanism examines the correctness of the student’s
answer. If the student’s answer is erroneous, the system records it and attempts to diagnose the nature of the
error. The system holds a student model, namely it keeps a profile of every student, and provides
individualized help concerning performance and error diagnosis for three languages. In addition, the errors
that originated from language confusion are a matter that is deeply examined.
1 INTRODUCTION
Learning several languages is becoming a
worldwide necessity. Especially, considering the fact
that some languages are not widely used, students
have to learn at least two foreign languages at
primary school, one of which is English (Kurata N.,
2010). Furthermore, there is an increasing interest in
the use of computer-assisted language instruction,
especially in cases where the language to be taught
is not the students’ mother tongue (Kunichike et al,
1998).
The design of Intelligent Tutoring Systems (ITS)
is founded on two fundamental assumptions about
learning (Ferreira and Atkinson, 2009). First,
individualized instruction by a competent tutor is far
superior to the classroom style because both the
content and the style of the instruction can be
continuously adapted to best meet the needs of the
situation. Secondly, students learn better in
situations which more closely approximate the
situations in which they will use their knowledge.
Furthermore, according to Ditcharoen et al (2010),
computer-assisted language learning (CALL) can
better serve the individual needs and is able to
capture and analyze the learner’s performance
during the learning process, and return feedback to
the learner. Hence, students have greater
opportunities to learn a foreign language, using an
ITS than being in a traditional classroom. However,
the need of multiple language learning is not
depicted in current scientific efforts, as the scientific
literature is orientated to single language learning.
However, facing reality in our own country gave the
authors the stimulus to conform to these
developments.
In view of the above, we have developed a
sophisticated tutoring system that has been based on
the architecture of Intelligent Tutoring Systems
(ITSs). The system combines interactivity and
adaptivity to individual students needs, along with
multiple language learning, that is English, French
and German learning. The resulting system is called
CAMELL and is an abbreviation of Computer-
Assisted Multilingual E-Language Learning.
CAMELL includes all the standard attributes of an
ITS and complies with its architecture, which
consists of the domain knowledge, the student
modeler, the advice generator and the user interface
(Wenger, 1987). A novelty of the system lies in the
multilingual component and in the error diagnosis
process, which is carried out through the languages.
The existence of three different languages is not
a way of separate learning in the same system. The
system provides estimation of the learner’s
509
Virvou M. and Troussas C..
CAMELL - Towards a Ubiquitous Multilingual e-Learning System.
DOI: 10.5220/0003470605090513
In Proceedings of the 3rd International Conference on Computer Supported Education (UeL-2011), pages 509-513
ISBN: 978-989-8425-50-8
Copyright
c
 2011 SCITEPRESS (Science and Technology Publications, Lda.)

proficiency in the domain as well as his/her
proneness to commit errors. The facility of
individualized error diagnosis is particularly
important for students, who can benefit from advice
tailored to their problems (Virvou et al, 2000).
2 RELATED WORK
Teaching languages through computer-assisted
approaches is quite significant and a wide range of
Figure 1: The CAMELL system and its basic components.
computer scientists has been attracted over the last
decades. After a thorough investigation in the related
scientific literature, we found a large amount of
works, concerning language systems, which have
large linguistic coverage, but limited error analysis
procedures or vice versa.
AutoTutor is a system, developed by Graesser et
al (2005), which simulates a human tutor by
promoting the conversation and provides feedback
to the learner, pumps him/her for more information,
gives hints, fills missing information with assertions,
identifies and corrects bad answers, answers
learner’s questions and summarizes answers.
Scooter the Tutor, developed by Baker et al (2006),
is a system which gives a gaming student
supplementary exercises focused on exactly the
material the student bypassed by gaming, and also
expresses negative emotion to gaming students
through an animated agent. Another computer-
assisted language learning system is rEcho,
developed by Zhou et al (2007), which can give
relevance feedbacks through anatomy animation and
is based on deliberate data trained recognition to
give error trend relevant feedbacks. SignMT was
implemented by Ditcharoen et al (2010) to translate
sentences/phrases from different sources in four
steps, which are word transformation, word
constraint, word addiction and word ordering.
Another computer-based program on second
language acquisition is Diglot Reader, which was
developed by Christensen et al (2007) and is used in
a way that students may read a native language text
with second language vocabulary and grammatical
structures increasingly embedded within the text.
TAGARELA is an individualized instruction
program, implemented by Amaral et al (2007),
which analyzes student input for different activities
and provides individual feedback. Finally, VIRGE,
developed by Katsionis and Virvou (2008), works as
asn adventure virtual reality game but it has
educational content as well and supports
personalized learning based on a student modeling
component.
However, after a thorough investigation in the
related scientific literature, we came up with the
result that teaching multiple languages through an
integrated tutoring system is an approach that was
not investigated before. For this reason, we
implemented the CAMELL prototype system, which
incorporates user modeling and error diagnosis
components, while teaching three different
languages, namely, English, German and French.
3 ARCHITECTURE OF THE
CAMELL
The architecture of CAMELL follows the main line
of Intelligent Tutoring Systems (ITS) architectures.
The major functional components of an ITS
architecture are the domain knowledge, the student
modeler, the advice generator and the user Interface
(Wenger, 1987). In this section, we will briefly
describe each one of these components, except the
student modeler which will be described in the next
section, so as to give in detail the effective operation
and interoperability of our system.
The domain knowledge of the CAMELL system
consists of three languages, which are English,
French and German. For each one of these languages
the system follows the same logical structure. This
structure includes five novice level lessons for
beginner students. The first lesson is the learning of
the alphabet of the corresponding language. The
alphabet is given both in capital letters and in
minuscules. The second lesson encompasses the
learning of months and days, along with their
pronunciation. The third lesson encompasses the
genders and the articles, so as to render the students
capable of mastering these subjects. The fourth
lesson describes in detail the personal and the
possessive pronouns. The final lesson familiarizes
students with the verbs “to be” and “have”, as main
verbs. An important issue considering these lessons
is that the there is a multiple-choice test for each one
of the three last lessons, so that the students get
evaluated and examined concerning their knowledge
and comprehension of the previous lessons. If the
students are not found to be adequately prepared to
CSEDU 2011 - 3rd International Conference on Computer Supported Education
510

go on to the next lesson, they have to study again the
relative theory. We used multiple-choice exercises to
evaluate their performance, as they provided us with
the ability to measure the students’ achievements.
Moreover, the user interface is very significant in
educational applications, because it can stimulate the
student’s interest in learning (Virvou et al, 2000).
The main goal of interaction between a human and a
machine at the user interface is the effective
operation and control of the machine, and the
feedback from the machine, which aids the operator
in making operational decisions. This goal is
adequately accomplished in our system. Finally, the
friendliness of the user interface of the CAMELL
system is achieved by using, colors, animations and
images, so that it can attract the student’s interest in
the subject. In section 7, we made an effort to break
up this complex application into simple, discrete
pieces that can be individually studied, so as to
advance the readability and reusability of our
system. Hence, we used UML, which offers a
standard way to visualize our system’s architectural
blueprints.
4 OVERVIEW OF THE SYSTEM
CAMELL is a multiple language learning system,
which incorporates the learning of three languages
and offers “communication” between them. Entering
the main page of the application, the student is asked
to choose the language that s/he wants to learn. By
choosing the language of his/her interest, the user
has to give his/her credentials in order to log in and
start the lessons.
After the successful log-in of the user, the
system follows two different approaches, according
to the gender of the student. Apart from the student’s
gender, the system holds records considering each
user’s interaction, such as when and at which lesson
a student has exited from the educational
application. Hence, when a student logs in the
system, s/he can resume the lesson that s/he was
taught in his/her last visit. Accordingly, a new user
will have to start from the first lesson and has no
right to go on to the other lessons, if s/he does not
accomplish it.
The student’s card, which is illustrated in Figure
2, can be checked in every lesson and for every
language of our application. It constitutes a user
model through which the students’ performance can
be monitored for all the languages and whenever the
student wants.
Figure 3 illustrates the categorization of the
errors in the final test. The student can be evaluated
and check where s/he is wrong and what type of
mistake s/he has made, after s/he had filled in the
gaps. The different colors indicate different type of
errors. The red color in the field means error in
articles or pronouns. The green color means a verb
mistake. The yellow color means a spelling mistake.
The blue color means confusion with the German
language, while the purple means confusion with the
English language. Finally, the grey color indicates
an unanswered question. In the same time, the
system shows grade of the student, along with the
exact number of the errors in each category.
Figure 2: Student’s card, which includes the student’s
profile.
Figure 3: Error diagnosis progress in the final test.
The “communication” of the three languages of
the system concerning the student’s performance is
quite noteworthy. Namely, the system can give
advice to the user, as far as his/her performance in
the other languages. This operation consists of a
significant component of adaptivity, as the system
can store each student’s performance in the database
and can give advice, when it is necessary.
CAMELL - Towards a Ubiquitous Multilingual e-Learning System
511

5 STUDENT MODELING
Student modeling, as the model of a learner,
represents the computer system’s belief about the
learner’s knowledge. It is generally used in
connection with applications computer-based
instructional systems. Student modeling is crucial
for an intelligent learning environment to be able to
adapt to the needs and knowledge of individual
students. Virvou et al (2000) support that the student
modeler is responsible for preserving the system’s
estimation of the learner’s proficiency in the domain
as well as his/her proneness to commit errors.
CAMELL constructs a student model, which
gives assistance to the learner, providing feedback or
interprets for his/her behavior. One significant
element is that before the student’s starting a
multiple-choice test in another language, the system
informs him/her about his/her performance in the
corresponding test of the lesson of the already taught
language and gives him/her advice concerning the
test s/he is about to do. Moreover, concerning the
final test, the student modeler checks the student’s
answer and in case of an error and it performs error
diagnosis. In this case, the system checks the
complexion of the error and acts in a way that it will
be described in the next section.
A matter of great importance is the existence of a
long term user model for each student. The system
includes also a form, which keeps information about
the student’s progress in the three languages, the
total grade in each one of the three languages and all
the results of the tests. Moreover, this form can be
presented to students so that they stay aware of their
advance of knowledge.
6 CONCLUSIONS
CAMELL is an educational application which
combines the attractiveness and user-friendliness
with individualized help that an ITS can provide. In
particular, the system incorporates the student
modeling component for each user and performs
error diagnosis. Moreover, the system keeps each
student’s error history in one language that is
already taught and then provides advice in the tests
of the other languages. In order to perform error
diagnosis, the system bears a detailed categorization
of common student’s mistakes. The error diagnosis
process of the CAMELL system is especially
focused on errors due to confusion of the other
languages of the system, if the student learns more
than on language at the same time. Furthermore,
apart from the friendliness of the user interface, our
system is oriented to offer adaptivity and dynamic
individualization to each user that interacts with the
application.
REFERENCES
Chapelle, C., 1998, Multimedia CALL: Lessons to be
learned from research on instructed SLA, In Language
Learning and Technology, 2 (1), pp. 21-39.
Kunichike, H., Takeuchi, A., Otsuki, S., 1998, Automatic
Generation of Questions for a Comprehension Test in
English Learning, In Proceedings of ED-MEDIA,
ED-TELECOM 98, World Conference on Education
Multimedia and Educational Telecommunications,
1, pp. 767-772.
Plass, J. L., 1998, Design and evaluation of the user
interface of foreign language multimedia software: A
cognitive approach, In Language Learning and
Technology, 2 (1), pp. 40-53.
Wenger, E., 1987, Artificial Intelligence and Tutoring
Systems.
Graesser, A. C., Chipman, P., Haynes, B. C., Olney, A.,
2005, Auto tutor: An intelligent tutoring system with
mixed-initiative dialogue, In IEEE Transactions on
Education, Volume 48, Issue 4, Pages 612-618.
Virvou, M, Maras, D, Tsiriga V., 2000, Student modelling
in an intelligent tutoring system for the passive voice
of english language, In Educational Technology and
Society, Volume 3, Issue 4, Pages 139-150.
Ferreira, A., Atkinson, J., 2009, Designing a feedback
component of an intelligent tutoring system for foreign
language, In Knowledge-Based Systems,Volume 22,
Issue 7, Pages 496-501.
Ditcharoen, N, Naruedomkul, K, Cercone, N., 2010,
SignMT: An alternative language learning tool, In
Computers and Education, Volume 55, Issue 1, Pages
118-130.
Baker, R. S. J. D., Corbett, A. T., Koedinger, K. R.,
Evenson, S., Roll, I., Wagner, A. Z., Naim, M.,
Raspat, J., Baker, D. J., Beck, J. E., 2006, Adapting to
when students game an intelligent tutoring system, In
Lecture Notes in Computer Science (including
subseries Lecture Notes in Artificial Intelligence and
Lecture Notes in Bioinformatics), Volume 4053
LNCS, Pages 392-401.
Katsionis, G., Virvou, M., 2008, Personalised e-learning
through an educational virtual reality game using Web
services, In Multimedia Tools and Applications,
Volume 39, Issue 1, Pages 47-71.
Zhou, W., Zheng, J., Lu, Q., Chiu, T., You, X., Ye, W.,
2007, A computer assisted language learning system
based on error trends grouping, In IEEE NLP-KE
2007 - Proceedings of International Conference on
Natural Language Processing and Knowledge
Engineering, Article number 4368040, Pages 256-261.
Christensen, E., Merrill, P., Yanchar, S., 2007, Second
language vocabulary acquisition using a diglot reader
CSEDU 2011 - 3rd International Conference on Computer Supported Education
512

or a computer-based drill and practice program, In
Computer Assisted Language Learning, Volume 20,
Issue 1, Pages 67-77.
Amaral, L., Meurers, D., 2007, Conceptualizing student
models for ICALL, In Lecture Notes in Computer
Science (including subseries Lecture Notes in
Artificial Intelligence and Lecture Notes in
Bioinformatics),Volume 4511 LNCS, Pages 340-344.
Kurata N., 2010, Opportunities for Foreign Language
Learning and Use within a Learner's Informal Social
Networks, In Mind, Culture, and Activity, Volume 17,
Issue 4, pages 382 – 396.
CAMELL - Towards a Ubiquitous Multilingual e-Learning System
513