A MODEL-DRIVEN AND EXTERNAL APPROACH

FOR LEARNING DESIGN UPON LEARNING

MANAGEMENT SYSTEMS

Aymen Abedmouleh, Pierre Laforcade, Lahcen Oubahssi and Christophe Choquet

Maine University – LIUM, IUT de Laval, 52 Rue des drs Calmette et Guérin, F-53020 Laval, Cedex 9, France

Keywords: Technology enhanced learning, Binding, Learning management system, Domain specific modeling.

Abstract: The research work presented is about helping teachers-designers in focusing on instructional design activi-

ties when using widely spread Learning Management System. We originally propose to focus on the implicit

LMS pedagogical language and to exploit it outside of the LMS by well-suited instructional design tools.

We concretely propose to follow a Domain Specific Modeling approach in order to formalize the LMS ab-

stract syntax (meta-model) and to use it as a basis for the elaboration and development of Visual Instruc-

tional Design Languages and dedicated tools. We also propose to extend LMSs with a communication API

for importing/exporting courses/scenarios from/to these graphical editors.

1 INTRODUCTION

Nowadays, Technology Enhanced Learning

environments like Learning Management Systems

(LMS) are widely spread and are not restricted for

intensive and distant learning uses. Most of

academic organizations provide teachers with some

LMSs and services for improving their face-to-face

courses by some additional activities.

Within the last decade, many languages and tools

have been designed in order to provide support to

the instructional design actors (Koper, 2006).

Notwithstanding these potential support, the design

and setting-up of distant learning situations have not

reached a mature and accessible level for providing

most of teachers with some user-friendly and

soundly all-in-one or automatic solutions. The

operationalization of learning scenarios is still an

issue.

We are interested by teachers-designers using

LMS within their academic organizations. They

directly use and handle platforms for setting up the

pedagogical situations they mentally designed

because of their weak instructional design culture

and weak help from their organization. They have to

understand the underlying “way of thinking and

designing” of these platforms, their implicit domain-

language. They also have to appropriate the various

screens and form-based interfaces, to abstract some

low-level details to think about the global design of

the courses they are setting up, etc. As a

consequence, LMS like Moodle are not spread and

used as they could do, essentially because most

teachers are not familiar with that implicit learning

design domain (Martinez-Ortiz et al., 2009a).

We aim to help such practitioners to better use

and apprehend the LMS implicit instructional design

language, and then help them to improve the design

of learning situations. We are then interested in

instructional design approaches and solutions that

could be either internal or external to the LMSs,

taking into account the binding and

operationalization of produced scenarios and in

relation to the LMS instructional design semantics.

2 EXISTING APPROACHES

Several approaches aim at facilitating the design of

courses by focusing on the specification of learning

scenarios and their binding into concrete LMSs. The

intervention of platform experts is no more

indispensable but these approaches require an

infrastructure for interacting with the platform and

for taking in charge the automatic creation and

configuration of the working spaces, as well as the

activity performance, starting from a formalized

description of the targeted learning situation. Such

398

Abedmouleh A., Laforcade P., Oubahssi L. and Choquet C..

A MODEL-DRIVEN AND EXTERNAL APPROACH FOR LEARNING DESIGN UPON LEARNING MANAGEMENT SYSTEMS .

DOI: 10.5220/0003339203980401

In Proceedings of the 3rd International Conference on Computer Supported Education (CSEDU-2011), pages 398-401

ISBN: 978-989-8425-49-2

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

approaches require a 'domain language' or

Educational Modeling Language (EML), allowing

modeling the learning activities, as well as a

'binding' technique to be machine-readable. These

approaches also require some techniques and tools to

support the 'operationalization' step consisting in

bridging the gap between the formalized and

machine-readable learning situations and their

concrete setting-up into a dedicated learning

environment.

The LAMS solution (Ghiglione et al, 2007)

consists in integrating a graphical internal editor to

some LMSs like MOODLE. It offers a user-friendly

interface to designers but produced scenarios or

courses are related to a specific runtime engine to

add to the LMS. This approach does not rely on the

LMS internal semantics. Other approaches focus on

the proposition of specific EMLs or Visual

Instructional Design Languages (VIDLs) with

'larger' semantics (Botturi et al., 2007) but they also

raise operationalization issues like the loss of

information. Excepting the all-in-one infrastructures

and EML/VIDL for practitioners, specified and built

together (e.g. the LDL language and the LDI runtime

infrastructure dedicated to play LDL scenarios

(Martel et al., 2006), none of the current

instructional design propositions concerns a direct

operationalization of practitioners-centered learning

scenarios on some LMS or direct transformations

towards equivalent scenarios conformed to some

LMS centered languages.

The COLLAGE proposition (Hernández-Leo et

al., 2006) is interesting because the collaborative

design patterns proposed to practitioners have been

specified and developed on top of the IMS-LD

standard: semantics about concepts/relations

transformations have been taken into account when

building the patterns; these patterns are so fully-

compatible with IMS-LD. The operationalization of

COLLAGE models then tackles the problem of

operationalizing IMS-LD models. Unfortunately,

existing LMSs are still not compatible with this

standard (Berggren et al., 2005).

Although CopperCore can be used as an IMS-LD

runtime engine, such complex tool is, as far as we

know, rarely used or integrated to LMSs. Moreover,

the scenarios specified by Collage, or other editor

dedicated to specific EMLs (IMS-LD, LDL, CPM,

etc.) do not focus on LMSs languages (ie. the LMSs

learning paradigms and features).

Also, most of research works that deal with the

exportation or transcription of learning scenarios

have highlighted the semantic learning design gap

that appears when considering learning scenarios

concepts and platforms features (Abdallah et al.,

2008, Caron et al., 2005). Such scenarios

transcriptions lead to some losses of information

from the source scenario or to some incomplete

informations into the platform transcription (lack of

sufficient information from the source model to

specify at the required level the platform elements).

This conceptual gap between two learning design

languages is inherent to the transformation process

when both languages have been elaborated with no

reciprocal relations.

3 FOCUSING ON THE LMS

SPECIFIC LANGUAGE

Current propositions rely on a same underlying idea

about evolving existent LMS by large add-ons

(editors or runtime engines) and new semantics in

order to integrate learning design standards or to

improve the design.

We do not aim to add new semantics to the

domain specific model embedded into the LMS. We

assume that each LMS is not pedagogically neutral

and that it embeds an implicit language for

describing the process of designing a learning

activity. Thus, our proposition is based on the

following ideas this language can be identified and

explicitly formalized in a computer-readable format;

this format can be used as a binding format for

various external tools which will focus on different

designing facets. LMSs have to be able to

import/export learning scenarios in conformance

with this language: current platforms have

notwithstanding to evolve in order to offer this new

functionality. From an LMS viewpoint, our

proposition is to add a similar 'import/export'

functionality like the SCORM one but with their

own language. We propose so a kind of new labeled

standard: self-compliance LMS. This will warrant e-

learning tools developers that they could exploit this

explicit language (that will have to be accessible

through an XML schema for example) for

communicating with the LMS. Operationalizing a

learning scenario from this LMS-centered viewpoint

will consist then in the importation of a learning

scenario formalized in conformance to this explicit

LMS language.

We also propose an original TEL-centered

Model-Driven Engineering and Domain-Specific-

Modeling (DSM) (Kelly et al., 2008) approach both

to identify/formalize the LMS language and to use it

as a basis for the elaboration of LMS-centered

A MODEL-DRIVEN AND EXTERNAL APPROACH FOR LEARNING DESIGN UPON LEARNING MANAGEMENT

SYSTEMS

399

VIDLs and their dedicated authoring-tools. From a

metamodeling viewpoint, every LMS language can

be considered as composed of an abstract syntax

(formalized as a metamodel and additional well-

formed rules), a concrete syntax (the machine-

readable textual notation that will be used for the

binding of learning scenarios), and some semantics

for both syntaxes.

The explicitation of LMSs languages allows the

specification of VIDLs/EMLs on top of them on

several layers of abstraction. This approach will

propose also a new opportunity to operationalize

learning scenarios. A first step for this approach is to

provide practitioners with some external learning

design editors based on the LMSs languages. It is

also important to provide practitioners with some

learning design editors dedicated to the VIDLs built

on top of the LMS language. Many VIDLs can be

proposed for a same LMS language. These LMS-

centered VIDLs have to be composed of the same

abstract syntax than the LMS language (same

domain meta-model), but have to propose a visual

notation (e.g. concrete syntax) in order to facilitate

thinking and communication for practitioners

(human-interpretable formalism). In contrast, the

dedicated editors of these VIDLs have to manage the

persistence of produced learning scenarios in the

machine-readable format of the considered LMS.

Our propositions are focusing on a DSM

approach that aims to offer a practical solution to

produce scenarios according to the semantics of the

LMS language. DSM tools will manage the binding

to the LMS machine-readable format. We propose to

use them in order to elaborate some LMS-centered

VIDLs and dedicated user-friendly editors based on

the meta-model of the identified LMS language.

We have experimented such DSM tools, the ones

from the Eclipse Modeling Project, able to specify

all these artifacts (domain meta-models, graphical

and textual notations, generation of dedicated

editors, etc.). These tools have been experimented

within several projects of different scopes and

following practitioners centered viewpoints as well

as TEL-centered ones (Laforcade , 2010).

Our LMS-centered solution guarantees that

future produced scenarios will be implemented on

the concerned LMS taking account the probability

that this solution may restrict the pedagogical

expressiveness of learning scenarios. But we assume

that the explicitation of the internal LMS language

will create the opportunity to build more

practitioners-directed but LMS-centered on top of

the initial LMS language.

4 A FIRST PROTOTYPE

For evaluating the potential of this approach, we have

initiated an action whose aim is to develop a simple

external editor based on the Moodle-centered

language and to study how to design and

operationalize a learning scenario with this editor.

The first step was to study Moodle itself from a

user centered viewpoint in order to identify the main

concepts of the implicit Moodle language. Then the

analysis of the various Moodle interfaces allowed us

to identify the language. We have refined this

language by analyzing the internal Moodle database.

Then we have capitalized this knowledge into a

specific metamodel.

Figure 1: Domain definition model.

On top of this metamodel we have developed a

very first visual editor. It aims to graphically ease the

specification of sections within a course. We used the

following DSM tools: EMF (main metamodel-

oriented framework) and GMF (graphical

framework). A full-generated prototype, as a plug-in

for Eclipse has then been generated by the DSM

tools in accordance with the domain and graphical

objectives. Concretely, this prototype is a simple

application providing a drawing space in which

graphical and inter-related course sections can be

modeled (Fig. 2).

This editor does not require computer skills: the

teacher/designer has to use the functionalities offered

by the tool palette to graphically specify a diagram

representing the course. He defines all the course

sections by creating them and filling some

information fields (activity name, summary, etc.). He

draws some links between sections for defining the

learning sequence. The scenario is serialized in an

XML machine-readable format.

For allowing the implementation of the scenarios

we have developed a block (e.g. a specific

functionality added to Moodle author environment)

which parses the XMI file and creates all the

concepts composing the Moodle course, according to

the scenario elements. It plays the role of a

communication bridge between the external editor

and Moodle. This block appears in the course screen

CSEDU 2011 - 3rd International Conference on Computer Supported Education

400

in design mode. The teacher has not anymore to deal

with the different interfaces previously needed for

specifying a course: he has just to choose the file to

import. After this importation step, the sections, title

and orders are directly set-up within the space course.

Figure 2: The external graphical editor.

5 ON GOING WORK

For going farther this first result, we actually work

on two directions: defining more complex VIDLs for

Moodle and evaluating the usability of the editors

built on them. We also aim to study at least one

another LMS, to repeat our proposals (e.g. defining

externals editors of pedagogical scenarios) in order

to evaluate the possibilities of interoperability

between two different technical frameworks, helped

by Model Driven Engineering techniques.

REFERENCES

Abdallah, F., Toffolon, C., and Warin, B., 2008. Models

transformation to implement a Project-Based Colla-

borative Learning (PBCL) Scenario: Moodle case

study. In 8th IEEE International Conference on Ad-

vanced Learning Technologies, Spain, July 2008.

Berggren, A., Burgos, D., Fontana, J. M., Hinkelman, D.,

Hung, V., Hursh, A., and Tielemans, G., 2005. Prac-

tical and Pedagogical Issues for Teacher Adoption of

IMS Learning Design Standards in Moodle LMS. In

Teacher Adoption of IMS Learning Design Standards

in Moodle LMS. Journal of Interactive Media in Edu-

cation, 2005.

Botturi, L., Todd-Stubbs, S., 2007. Handbook of Visual

Languages for Instructional Design: Theories and

Practices. Information Science Reference. ISBN-13:

978-1599047317.

Caron, P. A., Derycke, A., and Le Pallec, X., 2005. Brico-

lage and Model Driven Approach to design distant

course. In E learn 2005, world conference on E-

learning in corporate Government, HealthCare & high-

ter education.

Ghiglione, E., Dalziel, J., 2007. Design principles for

LAMS version 2 and the LAMS “Tools Contract”. In

Proceedings of the TenCompetence Conference Work-

shop. UPF - Barcelona, Spain.

Hernández-Leo, D., Villasclaras-Fernández, E. D., Jorrín-

Abellán, I. M., Asensio-Pérez, J. I., Dimitriadis, Y.,

Ruiz-Requies, I., and Rubia-Avi, B., 2006. Collage, a

Collaborative Learning Design Editor Based on Pat-

terns Special Issue on Learning Design, Educational

Technology & Society. 9(1), pp. 58-71.

Kelly, S., Tolvanen, J.-P., 2008. Domain-Specific Model-

ing. ISBN: 978-0-470-03666-2. Paperback. 427 pages.

Wiley-IEEE Computer Society Press.

Koper, R., 2006. Current Research in Learning Design.

Educational Technology & Society, 9 (1), 13-22.

Laforcade, P., 2010. A Domain-Specific Modeling ap-

proach for supporting the specification of Visual In-

structional Design Languages and the building of ded-

icated editors. Journal of Visual Languages & Compu-

ting, Volume 21, Issue 6, Pages 347-358.

Martel, C., Vignollet, L., Ferraris, C., David, J.-P., and

Lejeune, A., 2006. Modeling collaborative learning

activities on e-learning platforms, In: Proceedings of

the Sixth IEEE International Conference on Advanced

Learning Technologies, pp. 707-709.

Martinez-Ortiz, I., Sierra, J. L., and Fernández-Manjón,

B., 2009a. Enhancing IMS LD Units of Learning

Comprehension. In the 4th International Conference

on Internet and Web Applications and Services. May

24-28, 2009, Venice, Italy.

A MODEL-DRIVEN AND EXTERNAL APPROACH FOR LEARNING DESIGN UPON LEARNING MANAGEMENT

SYSTEMS

401