OPENING TEL SYSTEMS FOR TEACHERS
A Domain-specific Modeling & Model-driven Engineering Approach
El Amine Ouraiba, Christophe Choquet and Philippe Cottier
Maine University, IUT de Laval, 52 rue des Docteurs Calmette et Guérin, 53020 Laval Cedex 9, France
Keywords: Open Technology Enhanced Learning Systems, Instructional Design, Open Pedagogical scenario, Learning
Session Adaptation, Domain-Specific Modeling, Model-Driven Engineering.
Abstract: Despite their quality, few TEL systems are actually adopted in educational institutions. These educational
technologies have not always the necessary flexibility for use in real educational contexts that often
requiring the rapid adaptations to new and often unexpected events (Cottier et al., 2008). Indeed, TEL
environements should be designed as “open” in which the teacher himself is able to lead the adaptation and
reengineering of learning system at an abstract level. In our work, we consider that opening of pedagogical
scenario allows for the opening of TEL system. This article focuses on an approach based on the Domain-
Specific Modeling and Model-driven Engineering for supporting practitioner teachers in their activities
through the instructional design process. In order to verify our proposal we took Hop3x as experimentation
field. Our objective is to open this TEL system for its users by providing them a user-friendly editor which
allows the design and adaptation of learning sessions at a high-level of abstraction. We illustrate the
development process of Hop3x’s Domain-Specific Language and specific editor.
1 INTRODUCTION
Technology-enhanced learning (TEL) systems are
dedicated to making their users learn.
A TEL system
is a complex environment that mobilizes human
agents (learner, teacher) and artificial ones in
interactions conceived in order to improve the
quality of the human learning. The design of a TEL
system is a significant effort for a learning
institution. It is a complex process that is expensive
in time and also in technical and human means. In
this process the designer is involved to perform
several choices about pedagogy, technology and
interaction modalities. It must that the designed TEL
system could be adapted to the evolution of usage’s
context, and be configured according to the
evolution of users needs and activities. According to
(Rogalski, 2003) the teacher, in his/her activity,
should manage an "open dynamic environment".
Dynamic because the learning process evolves even
without teacher’s intervention, this is called a
spontaneous evolution. Open because the teacher
cannot predict the spontaneous evolution of learners
and the effect of his/her possible interventions
(Roditi, 2003).
(Henri et al., 2007) affirm that the learning
environment is a work in progress which is
improved session after session. This particularity
requires, according to (Cottier and El-Kechaï, 2009),
to think the design as a continuous and situated
process. However, the engineering’s classical
methods are characterized by their rigidity and
linearity. In one hand, after that the developed
system is implanted, the only evolution that can have
been is its maintenance. In other hand, for building
this system, the design process is performed in steps
that segment the time, the actors, and the works to
realize (Bourguin and Derycke, 2005). This brings
about a discontinuity between the design process and
the usage process (separation between designers and
users), and difficulties to fill the gap between the
scientific disciplines mobilized, in particular
between computer science and human and social
sciences (Bowker, Star and Turner, 1997),
(Bourguin and Derycke, 2005).
TEL systems should be designed as “open” (e.g.
systems where the teacher is able to lead the
adaptation and reengineering of learning system by
himslef). The question asked by engineering of TEL
systems is to “develop an adaptable and
reconfigurable TEL system by its users according to
312
Amine Ouraiba E., Choquet C. and Cottier P..
OPENING TEL SYSTEMS FOR TEACHERS - A Domain-specific Modeling & Model-driven Engineering Approach.
DOI: 10.5220/0003345803120317
In Proceedings of the 3rd International Conference on Computer Supported Education (CSEDU-2011), pages 312-317
ISBN: 978-989-8425-49-2
Copyright
c
 2011 SCITEPRESS (Science and Technology Publications, Lda.)

usage’s context evolution” rather than to “develop a
TEL system according to a given specification”. In
the open TEL systems engineering, the decisions
and practices of instructional designer are
fundamental (Cottier and El-Kechaï, 2009).
We consider that the pedagogical scenario is a
relevant and strong model for TEL systems
engineering and that the opening of learning
scenario allows the opening of TEL system. The
scenario is formalized with the help of an EML
(Educational Modeling Langage) (Koper, 2001),
defined by a specific meta-model which is itself
linked by conformity relations with the scenario.
Within this framework, two approaches exist in the
research field for the scenario based TEL systems
engineering (Choquet, 2007) : (1) interpretative
approach, where an existing EML (such as (IMS
LD, 2003)) is proposed to designer for specifying
scenarios; (2) constructive approach, where the
designer, generally helped by modeling specialists,
build the meta-model which describes their domain-
specific EML and use it for specifying scenarios.
Our work falls under this last approach. By the
use of Domain-Specific Modeling (DSM) and
Model-driven Engineering (MDE) paradigm we
want to surpass the difficulties a teacher can
encounter when using generic EMLs and existing
editors (Ouraiba et al., 2010). This, by defining the
Domain-Specific Language (DSL) of teacher and
developing accordingly a dedicated editor. The next
section of this article presents the DSM/MDE
paradigm how we instantiates it to support teacher
both at design and run time. Thus, we took Hop3x as
experimentation field of our approach. In section 3,
we detail the work realized on the Hop3x TEL
system for opening it where we illustrate our first
results of the development process of Hop3x’s
Domain-Specific Language (DSL) and specific
editor thanks to EMF tooling. We conclude by
discussing the benefits of our approach and our
future works.
2 DSM / MDE FOR SUPPORTING
TEACHER
Model-Driven Engineering (MDE) is basically a
software development approach where the code is
produced by some transformations and combinations
of models of the required artifact. It is an
enhancement of the Model-Driven Architecture
(MDA) approach, initially proposed by Object
Management Group (MDA) in 2001 (the MDA
document Guide (OMG, 2006) provides an overview
and definitions of the used concepts) to provide a
solution to the problem of software technologies
continual emergence that forces companies to adapt
their software systems every time a new “hot”
technology appears (adaptability problem).
The first principle is to use modeling and models
to develop software systems. The second principle is
the separation of the enterprise functionalities of an
information system from the implementation of
those functionalities on specific technological
platforms (EJB, CORBA, and so on). The abstracted
specification of the system becomes the main asset
in software development: many implementations
using concrete technologies may be derived from the
specification. It is model-driven while “it provides a
means for using models to direct the course of
understanding, design, construction, deployment,
operation, maintenance and modification” (OMG,
2006). The software development process is based
on automatic or semi-automatic transformations
between models, from abstracted, domain centered
and generally informal models (Computation
Independent Model – CIM) to specific and platform
dependent ones (Platform Specific Model – PSM).
Based on this initial proposal, the Domain
Specific Modeling approach (DSM) was defined (1)
to reduce the complexity of the transformations and
the semantic losses they generate, and (2) to enhance
the level of abstraction of the software specification
(Kelly and Tolvanen, 2008). The principle here is to
develop a Domain Specific Language (DSL),
tailored for specifying software which instruments a
specific activity in a specific context. This DSL has
to be formal but its meta-model reflects the domain
of the users: the modeling vocabulary used is the
domain one. Then, code generators could be
developed for directly transform models expressed
with a DSL into a specific technological platform
framework.
Particularly, we propose to adopt a DSM/MDE
approach for allowing the teacher to assume his/her
designer role, and the learning session actors to
adapt the open pedagogical scenario dynamically.
We consider that a pedagogical scenario, for being
really designed and manipulated by a teacher, has to
be considered as a domain specific model, expressed
with a DSEML (Domain-Specific Educational
Modeling Language) situated in his/her teaching
context and rooted in his/her practices. In such a
paradigm, MDE techniques have to support the
transformation of the scenario from domain specific
representation to operationalized one, both at the
design phase to support the operationalization and at
runtime to support the dynamic adaptation (Ouraiba
OPENING TEL SYSTEMS FOR TEACHERS - A Domain-specific Modeling & Model-driven Engineering Approach
313

et al., 2010).
The following figure 1 illustrates how we
instantiates DSM/MDE paradigm to support teacher
both at design and run time.
Figure 1: OMG layers view of the open pedagogical
scenarios engineering.
3 DSM/MDE FOR OPENING
HOP3X TEL ENVIRONMENT
FOR TEACHERS
3.1 Hop3x TEL System
Hop3x is a practical works TEL environment
developed for learning and teaching object-oriented
programming languages, like: C, Ruby, and Java.
We will focus here on Java programming. On one
hand, Hop3x is structured around a specific Java
editor/compiler where the learner has to solve
programming exercises. In the other hand, it allows
a tutor to intervene remotely and at runtime by
providing help and recommendation to learners
thanks to feedback system which provides high level
information about learners' productions and tasks by
the way of indicators calculated from tracks with
DCL4UTL language (Pham Thi Ngoc et al., 2009).
Hop3x is operating-systems independent and its
architecture is composed of three applications:
 Hop3x-Learner Interface, which allows
learners to write, edit, compile and run code
and program. It also allows them to ask for
help from the teacher via a synchronous
communication tool.
 Hop3x-Server, which collects interaction
tracks of learners and save them as Hop3x
events. It allows real-time calculation of
indicators.
 Hop3x-Tutor Interface, which is a monitoring
tool for tutors. It allows them to manage a
group of learners in a situation of distance but
synchronous practical work.
After their authentication, learners can use their
working interface. Before trying to solve exercises,
learners must read the what-to-do and how-to-do
instructions. The set of questions is presented as a
sequence that is not static, in a way where learners
can browse the list of questions and choose the order
that they want to follow. Answering questions
requires writing a Java-code. Generally, before
obtaining a correct and executable Java-code,
learners perform a series of updates on it (writing,
editing and deletion), sometimes compilations and
runs. If a learner finds a difficulty in his/her activity
he/she can request help from tutor via the audio
communication functionality.
The tutor, after authentication, can use the
relevant interface which is composed of a set of
functionalities developed in order to help him in
sessions control and learners monitoring during their
activities. These functionalities allow to: know
which learners are connected and which are not;
observe at real-time the Java-code written by each
learner (in fact, a mirror of the learner’s interface);
replay with a selected speed the different stages a
learner’s session; have a look on each compilations
and executions results for a given learner; make
tutoring interventions via audio or textual (proactive
messages) modalities; responds orally or textually to
a learner’s call (reactive message); see the history of
the tutoring interventions, read the content of
questions and instructions; consult specific and
transversal indicators calculated at real time.
Each tutoring intervention is characterized by its
manner, its modality, its category and its strategy. A
proactive intervention could be motivated by the
Java-code written quality and/or thank to indicators
characterizing the learners activities. A reactive
intervention could be fed also by indicators. These
indicators are calculated using the language
DCL4UTL (Pham Thi Ngoc et al., 2009). The
dimensions of tutoring interventions are classified
into four categories (Després, 2001): didactical
support about the taught content; methodological
support about learning organization; technical
support about resources provided to perform
activities; and motivation about the psychological
and emotional state. Thus, the intervention strategy
CSEDU 2011 - 3rd International Conference on Computer Supported Education
314

corresponds to the action that can be performed by
tutor such as: making a course reminding, modifying
a question, recommending the consultation of an
additional (external) resource, encouraging learner
to document his/her Java-code, explaining in order
to help learner to have some reflection, etc.
3.2 Hop3x Learning Sessions
The sessions we have chosen to highlight here are
parts of a course entitled “Object-oriented
programming and Java”. The learners involved are
undergraduate learners. These learners are novices in
Java programming but they had, during the
preceding term, an introductory course entitled
“Introduction to Object-oriented Programming”.
They attended lectures and tutorials about the basics
notions and concepts of object-oriented
programming such as class, object, instance,
message passing, inheritance, encapsulation,
overriding, overwriting and polymorphism. Then
they use the Hop3x TEL system in order to
implement these concepts in situation.
In the version used here, Hop3x-server can only
run learning sessions that are provided in the format
of a XML file. Thus, to design a Hop3x learning
session, designer (which is a teacher) writes
manually a XML file in which he/she defines its
structure (help by a template). He/she identifies
firstly the basic elements that compose the
mandatory layer of the session’s structure (section
4), these elements are: session’s characteristics
(name, programming language, pedagogical
objective, time of its start and end), actors who can
be involved (learners, tutors and groups), the
instructions to be respected by learners, and some
elements that can be used such as the necessary
resources. Secondly, he/she defines different
learning sequences (scenarios) that he/she can
anticipate. The set of these sequences defines the
foreseen contextual layer in the session’s structure.
A learning scenario is composed of a set of
questions characterized by an ID, a content, a set of
indicators and three types of tasks: required (what
the learner has to do), optional (what the learner has
of doing to think) and prohibited (what the learner
cannot do).
3.3 Development Process of Hop3x
DSEML and Specific Editor
The Hop3x system is developed initially for
executing linearly learning sessions, without
deviation and dynamic adaptations at runtime of
predictive learning scenarios. Our ultimate objective
is to “open” more this TEL system which remains
rather closed in its current version. We think that the
opening of a TEL system can be realized by its
reengineering in order to have a new version which
can allows its users (especially teachers) to design
and adapt dynamically learning sessions according
to the current context. We want indeed to enable
teachers to participate actively in the adaptation and
reengineering activities.
We aim by our work to provide the necessary
tools and conceptual means to teachers who use
Hop3x to enable them to design and adapt learning
sessions at a high-level of abstraction without
needing to create XML files by hand or with the
help of a generic tool (such as Reload Editor
(Reload, 2004)), we adopt the DSM/MDE approach
to help teachers to define their business language
(DSL) for developing accordingly a specific editor
more user-friendly. For this, we followed a
pragmatic methodology that starts from an existing
situation that is of legacy Hop3x TEL system, and
then try to propose improvement solutions to make it
more open for users who haven’t much technical
knowledge. First of all, through a first process that is
presented in the rest of this article we want to
develop knowledge about the relevance of the
implementation of specific tools, which are based on
the users business, without taking into account the
dynamic adaptation aspects.
This first process of our methodology lasts 4
steps (see Figure 2). In the first step, we investigated
the semantic of Hop3x: we collected and analyzed
the use cases of the actual version of Hop3x for
extracting domain specific concepts and rules
(Sections 3.1 and 3.2). Then based on this, we
specified the metamodel that describes the Hop3x’s
domain specific language. This metamodel
formalizes the semantic of Hop3x field by
specifying the meaning of each concept and how it
can be used according to domain‘s rules and
respecting constraints. In the third step, a Hop3x-
specific editor was generated from the DSL
metamodel. This editor makes available as
specification tools the concepts and rules that are
handled usually in the Hop3x practices. Finally, a
teacher could use this editor for designing the
practical works sessions at an abstract level.
Figure 2: Development process of Hop3x DSL and
specific editor.
OPENING TEL SYSTEMS FOR TEACHERS - A Domain-specific Modeling & Model-driven Engineering Approach
315

3.4 Use of the EMF Tooling
The Eclipse Modeling Projects (EMP, 2008)
provides a unified set of modelling frameworks,
tooling, and standard implementations, such as EMF
(Eclipse Modeling Framework), GMF (Graphical
Modeling Framework) and ATL (ATLAS
Transformation Language). In the following, we use
the EMF because it facilitates code generation for
building tools and other applications based on a
structured metamodel (Steinberg et al., 2008). Our
objective was to specify a metamodel which
describes the Hop3x’s DSL, and then generating
from this metamodel the code of the editor thanks to
tools provided in EMF. Indeed, this metamodel is an
ECORE model where ECORE is the MOF-like
meta-meta-model in EMF. Figure 3 illustrates this
metamodel in the class-diagram-oriented view
proposed by the ECORE graphical internal editor of
EMF.
Figure 3: Hop3x’s DSL metamodel.
A first version of the editor has been generated
automatically from Hop3x’s DSL metamodel thanks
to the EMF tooling. This editor provides a tree-view
of the models which are namely the Hop3x learning
sessions (see Figure 4). By using this editor the
teachers who want to use Hop3x can design the
practical works sessions at an abstract level
compared to the manual creation of XML files as it
is the case currently.
Figure 4: Example of a Hop3x learning session designed
by the specific editor.
Finally, thanks to this editor, designer can
generate learning sessions in the XML format
required by Hop3x system. The following figure 5
shows an example of a Hop3x learning session
generated as an XML file after its design by the
specific editor.
Figure 5: XML file of a Hop3x learning session designed
by the specific editor.
Recently, we have conducted a testing
experimentation of the specific editor with the
students who preparing the “Professional License in
the Design and Realization of Multimedia Services
and Products”. These 22 students had particularly
lessons related to learning design. As work they had
to describe a practical learning session of Hop3x
using two editors separately, the first one is generic
(Reload Editor (Reload, 2004) which implements
IMS LD (IMS LD, 2003)) and the second one is the
specific editor which we have developed based on
Hop3x’s DSEML (see Figure 4). Our goal was
simply to verify which editor was intuitive enough
to enable the autonomy of its user. Beyond this
testing, we have noted the interests of "putting in the
hands of users" a specific editor, freed from the
conceptual and technical barriers of learning
session’s representation.
CSEDU 2011 - 3rd International Conference on Computer Supported Education
316

4 CONCLUSIONS
The pedagogical scenario can be considered among
the strong models for investigating the engineering
of open TEL systems. It is relevant to open TEL
system through the opening of scenario. Our work
falls under the constructive approch of instructional
design of open scenarios where we have adopted the
DSM/MDE paradigm. To verify our proposal we
took Hop3x as experimentation field. The ultimate
objective is to further open this TEL system for its
users. For this, we follow a pragmatic methodology
that spans on two processes. In the first one,
presented in this article, we deal with the potential of
specific tools of instructional design. While in the
second one, we plan to investigate the adaptation
possibilities by specific tools. Indeed, although our
methodology spends more time and effort, a first
benefit is that teachers have to develop a reflexive
analysis on their practices and what they could do
with the TEL system.
Using the developed Hop3x-specific editor we
are currently conducting interviews with Hop3x’s
users in order to promote the expression of the
adaptation requirements of learning sessions and the
openness needs of TEL system. The information
gathered from these interviews will help us to define
another version of metamodel which will be more
optimized in order to: (1) guide us in the perfection
of Hop3x’s functionalities for transforming it into a
TEL system more open, and (2) to develop a
graphical editor dedicated to the design and
adaptation of learning sessions at a high-level of
abstraction. We have planned to use the GMF in a
second time to add a graphical layer on top of EMF.
ACKNOWLEDGEMENTS
Authors acknowledge the designers and users of
Hop3x TEL Environment for their help and
information about their habitual practices.
REFERENCES
Bourguin, G., Derycke, A., 2005. Systèmes Interactifs en
Co-évolution Réflexions sur les apports de la Théorie
de l’Activité au support des Pratiques Collectives
Distribuées. Revue d’Interaction Homme-Machine
Vol 6 N°1.
Bowkers, G., Leigh Star, S., Turner, W., Gasser, L., 1997.
Social science, technical systems and cooperative
work: beyond the great divide. Lawrence Erlbaum
Associates, “Computer, cognition and work” series,
470 p.
Choquet, C., 2007. Ingénierie et réingénierie des EIAH-
L’approche REDiM. Habilitation à diriger les
recherches en informatique, Université du Maine.
Cottier, P., Choquet, C., Tchounikine, P. 2008. Repenser
l’ingénierie des EIAH pour des enseignants
concepteurs. In: Usages, usagers et compétences
informationnelles au XXIème siècle, Jérôme Dinet
(ed.), p.159-193. Édité par Hermes Lavoisier, ISBN
978-2-7462-2193-2.
Cottier, P., EL-Kechaï, H., 2009. L'usager concepteur en
situation: Conception collective d'un livret
électronique d'apprentissage (LÉA).In: Ingénierie des
systèmes d'information, ISSN1633-1311. vol.14, no3.
p162
Després, C., 2001. Modélisation et conception d'un
environnement de suivi pédagogique synchrone
d'activités d'apprentissage à distance, Phd Thesis,
Université du Maine, 286p
Eclipse EMP, 2008. Eclipse Modeling Projects.
http://www.eclipse.org/modeling/. Retrieved from
Octobre 2008.
Henri, F., Compte, C., Charlier, B., 2007. La scénarisation
pédagogique dans tous ses débats. Revue
internationale des technologies en pédagogie
universitaire, 4(2).
IMS Global Learning Consortium, “IMS Learning Design
Specification, Version 1 - Final Specification”, 2003,
available at: http://www.imsglobal.org/learningdesign/
Kelly, S., Tolvanen, J., 2008. Domain-specific modeling:
enabling full code generation. 448 pages. Wiley-IEEE
Computer Society Press. ISBN: 978-0-470-03666-2.
Koper, R., 2001. Modeling units of study from a
pedagogical perspective: the pedagogical meta-model
behind EML, Open University of the Netherlands.
OMG, 2006. MDA specification guide. Version 1.0.1.
Report—omg/03-06-01.
Ouraiba, E. A., Choquet, C., Cottier, P., Després, C.,
Jacoboni, P., 2010. Engineering of open learning
scenarios: the case of Hop3x learning scenarios.
Proceedings of IEEE ICALT’10, Tunisia, pp. 264-268.
Pham Thi Ngoc, D., Iksal, S., Choquet, C., Klinger, E.,
2009. UTL-CL:A Declarative Calculation Language
Proposal for a Learning Tracks Analysis Process.
Proceeding of the 9th IEEE International Conference
on Advanced Learning Technologies (ICALT'09),
Riga, Latvia, July, pp.681-685.
Roditi, É., 2003. Régularité et variabilité des pratiques
ordinaires d’enseignement. Le cas de la multiplication
des nombres décimaux en sixième, Recherches en
Didactique des Mathématique. 23 (2).
Rogalski, J., 2003.Y a-t-il un pilote dans la classe ? Une
analyse de l’activité de l’enseignant comme gestion
d’un environnement dynamique ouvert. Recherches en
Didactique des Mathématiques. 23(3)342-348.
Steinberg, D., Budinsky, F., Paternostro, M., 2008. EMF:
Eclipse Modeling Framework, Second Edition.
Publisher: Addison Wesley Professional.
The Reload Project, The University of Bolton, The
University of Strathclyde and JISC, 2004. Available
at: http://www.reload.ac.uk/editor.html
OPENING TEL SYSTEMS FOR TEACHERS - A Domain-specific Modeling & Model-driven Engineering Approach
317