A FLEXIBLE INFRASTRUCTURE FOR P-LEARNING: A FIRST
APPLICATION IN THE FIELD OF PROFESSIONAL TRAINING
Alain Derycke and Vincent Chevrin
Laboratory LIFL/TRIGONE, University of Sciences and Technologies of Lille, Bat 6, Villeneuve d’Ascq, France
Keywords: Pervasive-Learning, Context-Aware environments, Ubiquitous Computing, Dynamic Adaptation of
Services, Multi-channels accesses.
Abstract: With the availability of nomadic computing, and its new interaction user devices connected through wireless
networks, it is obvious that the traditional way of delivering e-Learning will be changed. This paper is
focused on a new mode called pervasive learning which relies on the potential of new IT infrastructures able
to provide dynamic adaptations of information contents and services according to various contexts. Using
our previous experiences in the design and implementation of multi-channel accesses to services (mobile-
Commerce or e-Learning) we are designing a new infrastructure, based on a Multi-Agent Systems, which
satisfies our requirements for future p-Learning systems. Its potential is illustrated through a dedicated
scenario of uses drawn from needs founded in the field of learning on demand, in the framework of a shop,
contextualized for several seller situations and professional activities. The dedicated system, called a
Personal Training Assistant, is supported, in interaction with a Smartspace, through our infrastructure.
1 INTRODUCTION
With the rapid dissemination and uses of mobile
wireless devices, there is an opportunity to enlarge,
or to modify, the scope and the nature of the
traditional e-Learning modes of education. This is
particularly required by the organizations that are
rapidly evolving, under the pressure of their users
and of their markets, whatever the users are students,
clients, citizens or members of a community. Recent
surveys, such as those conducted by (Woodill,
2006), show also that new socio-technological
approaches of the Web, the WEB2.0 umbrella; will
have a great impact on the way we will consider
exchanges of information and co-production of
knowledge in the future. So we can assume, from a
research perspective, that the features of the e-
Learning will be more open, but also more complex,
in order to incorporate all these social and
technological changes, into a co-evolving process,
occurring between the technological pushes and the
real needs of the stakeholders.
Using our previous experiences in the design,
development, deployment of e-Learning services
and technological platforms, both from the research
or real operational perspectives, and also our
experiences in research on multi-channel computer
applications in the field of e-Commerce and mobile-
Commerce, we have start a new research program
focused on the design and experimentation of new
infrastructure for pervasive learning mode. This is
done in the framework of an important collaboration
in a large regional consortium, between Universities
and several big companies from the retail industry,
and from the e-Commerce and the direct marketing
fields, sets to forecast and master evolutions of these
fields. Our present research project is deeply rooted
in the new domains of computer sciences whatever
they are called “Ubiquitous Computing”, “Context-
Aware Computing”, “Pervasive Computing” or even
“Ambient Intelligence”. But it takes also seriously in
consideration the particular needs from these
companies, and their application domains, for the
continuous training of their employees and, if
possible jointly, of their clients and affiliates. This
will be illustrated by a specific scenario of use,
simplified here, that we have kept for presentation in
this paper, among several others that we are
investigating actually.
215
Derycke A. and Chevrin V. (2007).
A FLEXIBLE INFRASTRUCTURE FOR P-LEARNING: A FIRST APPLICATION IN THE FIELD OF PROFESSIONAL TRAINING.
In Proceedings of the Ninth International Conference on Enterprise Information Systems - HCI, pages 215-222
DOI: 10.5220/0002361902150222
Copyright
c
SciTePress
2 FROM E-LEARNING TO
P-LEARNING
2.1 Mobile Learning is more than
e-Learning Adaptation to Mobile
Phone
At a first look it seems that mobile-Learning (or m-
Learning) is just an adaptation of the e-Learning
systems to accesses through wireless networks,
increasing the accessibility (from the learner
location) and the “reachability” (to the learner
location). There is already a strong research interest
in the world for this new domain. It appears that m-
learning is, first, not a mode phenomenon, and
second, that it goes beyond traditional e-Learning
systems. Several research projects (Attewell, 2005)
have shown that in spite of the current limitation of
the mobile devices put in the users’ hands (i.e. a
basic GSM phone), it is possible to envision new
learning activities, less focused on rich document
interactions, but more communication oriented, in
the spirit of the CSCL mode of education. The
novelty in uses of wireless networks and mobile user
devices is the possibility to take into account the
singular context of the learner/user, for example
depending of his/her location. m-Learning is
learning that can take place anytime, anywhere with
the help of a mobile computer device. The device
must be capable of presenting learning content, in a
context sensitive way, and providing wireless two-
way communication between teacher(s) and
student(s). This is why m-Learning is different from
e-Learning due to its ubiquitous nature. However
previous researches have shown that to be effective
the user device must satisfy some requirements:
Highly portable, Individual, Unobtrusive, Available
anywhere, Adaptable to the context of learning and
the learner’s evolving skills and knowledge,
Persistent, Useful, Intuitive to use by people with no
previous experience of the technology (Sharples,
2003).
The researches on m-Learning have taken two
directions:
- The design of systems based on wireless LAN
(WiFi or Bluetooth) that organize the learning
physical space, classroom, theatre, campus, in order
to supplement the mobile user devices, in general
Personal Digital Assistant, with fixed devices
located close to the users. This is a case of what it is
called “SmartSpace”, where there is an automatic
detection of the user location and of the proximity
services. For example the PDA can be used in
conjunction with a large intelligent electronic
whiteboard in a theatre during a lecture.
- The design of specific portals, for regular
Learning Management Systems, LMS, that are able
to support mobile communication and devices on
GSM type of networks in Europe. This means
gateways for SMS supports, sometimes the support
of the WAP, or I-Mode standards, for browsing
information. But until now there are no true multi-
channel accesses to these LMS, in the way already
explored in the field of m-Commerce. Even worse
we have already demonstrated that Learning Objects
standards, such as SCORM compliant ones, can
limit the potential of a dynamic adaptation of the
learning contents to the diversity of channels (X3,
2006). Similarly, the potential of the multimodality
is not used in m-Learning, for example using the
potential of coupling voice and WAP interaction,
thanks to standards such as VoiceXML and X+V,
and the proposal for a true multimodal and device
independence Web in the future, by the W3C
consortium.
2.2 Emergence of the Pervasive
Learning Mode
In the last years a new concept has appeared to
translate the potential of Ubiquitous or Pervasive
Computing in education. This new way to use
technologies to support the learning processes has
been called either “Ubiquitous-Learning” (or u-
Learning) (Jones, Jo, 2004), or Pervasive-Learning.
(Keil-Slawick et al., 2005). In this proposal we have
adopted p-Learning to name this new field of
research. As for m-Learning, most of the authors
mention that p-Learning goes farther than uses of
new technologies provided by recent research in
pervasive computing to support e-Learning
traditional views, but that it enlarges also the view of
the learning process itself. For example for
(Bomsdorf, 2005): “Ubiquitous learning is the next
step in performing e-Learning…. Furthermore, it
enables seamless combination
of virtual
environments and physical space”. And she gives
the main characteristics of u-Learning in terms of:
Permanency, Accessibility, Immediacy,
Interactivity, Situating of Instructional Activities,
and Adaptability. It means than context sensitive
learning environments, learning material semantics,
are not necessary locked to the geographical location
of the learner, but depend of other kinds of contexts
such as his/her present role into the professional
activities and work, the history of the past learning
phases, the technological contexts (characteristic of
the present channels that can be used) and, in the
future, even the affective state of the learner. In
ICEIS 2007 - International Conference on Enterprise Information Systems
216
accordance with Bomsdorf, we state that the central
problem is about dynamic adaptation of the learning
materials and activities proposed for several kinds of
contexts. We also share the reflection done by
(Syvänen et al., 2005) when they said “The
combination of context and adaptivity…can also
prioritize communication channels
, encourage
cooperation, and adjust information modalities”.
It is the challenges that we want to support into the
open intermediation infrastructure we are designing
in our present project. And for that purpose we will
reuse also previous researches that we have
conducted in the field of multi-channel e-Commerce
(Chevrin et al., 2005a).
3 NEW NEEDS AND NEW
LEARNING SCENARIOS FOR
EVOLVING ORGANISATIONS
3.1 Some General Hypotheses for the
Design of Future p-Learning
Infrastructures
Just to summarize we want to put emphasis on some
of our main hypotheses, which result of our previous
investigations and experiences in the field of e-
Learning, or in the field of the mobile-Commerce:
H1: New learning modes, in this case, the p-
Learning mode, and new organization approaches
will appear that enlarge or modify the traditional e-
Learning mode. This is especially true due to the
ubiquitous nature of the interaction with the learner
on the one hand, and in the other hand to the new
needs of the organization for the training of both,
and jointly, of their employees, clients, providers
and partners in accordance with the related
organizational workflows. There is also a
tremendous potential for some learning
environments and contents that are more or less
auto-produced by the learners themselves, in
collaboration with others pedagogical agents (human
or synthetic one). This leaves place for a better
involvement and an active participation of the
person to his/her learning and collaborating process;
H2: There are various “contexts” that need to be
considered for a real p-Learning system, some are
independent of the domain of application such as
technological, physical contexts, and some are
specific to the learning process and to the present
status of the learner (worker or client). An effort
must be put on the way to “contextualize” a
pedagogical scenario and to support an easy
development of the deployment of the infrastructure
needed for these p-Learning settings;
H3: The first generation of technological e-
Learning platforms, the LMS, are not suitable for the
development of p-Learning, Learning on Demand
across the value-chains, and in the context of future
agile and virtual organizations. The future will be a
collection of dedicated learning and other e-
Services, some provided outside the concerned
organization, similar to the Application Service
Provider model;
H4: The first investigations about mobile-
Learning have shown that the mobile phone user
platform can be useful if proper adaptation of both
the learning material and of the learning activities is
done. The rise of the Smartphone, supporting
universal roaming (from GSM to Wifi and local
Internet accesses for example) and different
communication medium, from Voice, Podcasting,
WAP, Web 2.0, Rich Media, to Digital TV, should
be a booster for the new learning environments;
H5: Importance of the dynamic adaptations of
materials, e-Services and activities for learning, and
Contributions
Situations, Contexts
Equipment
designers
Products data
sheets
workflow
Activity
E-Learning
services
Channels
information
User devices
informations
Ont ol ogi ca l
Model Domai n
Kno wl edge
Hyper medi a
Document s
Shop
context ual i sed
documents
Rol e
context ual i sed
documents &
Cont ext uaki sed
e- Learnni g
servi ces
Mul t i èc ha nne l s
accesses
Concret e user
i nte rf aces
Shop stocks
& marketing
Levels of
expertises
Shop local
parameters
-Initial
training
-Recurrent
training
-In office
-In the shop
-In front of
the client(s)
User profiles
e-services
available
-Local or
remote?
-Wifi or
mobile
phone
-Earphone or
not
-Smartspace
-Geolocalisat
ion
-Type of
user
devices?
-Colocation
or not
-Multiusers
or not (SDG)
Figure 1: Chain of contextualized transformation for the scenario Personal Training Assistant.
A FLEXIBLE INFRASTRUCTURE FOR P-LEARNING: A FIRST APPLICATION IN THE FIELD OF
PROFESSIONAL TRAINING
217
of the delivery systems and their channels or
modalities mobilized for the user interaction, in a
great diversity of contexts and of rules and strategies
of transformation, that maintain nevertheless the
pedagogical and didactical intentions.
3.2 An Example of Rich Generic
Scenario with the Associated
Transformation Chain
In the framework of our present project to design a
p-Learning infrastructure we have partnerships with
companies in the field of retail, ranging from small
retail shops to hypermarkets and huge online stores.
The following generic scenario is extracted from the
needs in matter of p-Learning coming from one of
these partners. However it will be simplified here for
reasons of clarity and of confidentiality.
The main aim is to develop a Personal Training
Assistant (PTA) for sellers working in the field of
Hifi/Video equipments. This domain is becoming
complex due to the rapid changes in the underlying
technologies with the appearance of High Definition
TV, sophisticated home video, media-centers,
Terrestrial and satellites digital TV… So an effort
for continuous education of the sellers is needed in
order to maintain quality of the relations with the
clients and his/her efficiency. The main idea is to
use hand held computer, which can be used for
several others purposes also (stocks management for
example), in order to support both the learning and
the coaching of the seller/learner in various contexts.
It could be for example a Personal Assistant
connected to the information infrastructure of the
shop. Because the development of such a training
system is costly, it is envisioned that the cost will be
shared by a cooperation across, not only all the
shops of the distributor, but also across the value
chain. This means that the learning contents and
activities start from general and more abstract levels,
not contextualized, are transformed through a chain
of transformation, in order to generated, mostly at
run-time, a dedicated, contextualized support for a
particular seller/learner interaction situation. It must
be noted that the whole system has also intents to be
adapted automatically to the training of the potential
clients, in order to facilitate her/his choice of
particular set of equipments, or for after a purchase
for its operations.
Figure 1 shows that many transformations must
be supported in order to deliver a pertinent learning
or coaching support to a particular user that can be
mobile, across the shop and connected through
wireless networks. Various situations (or contexts)
can be handled by the system depending of the
place, work activities, presence of the potential
clients or not, etc. Several sources on contribution
feed the chain with knowledge and contents. It must
be noted that the other e-Learning services such as
tutoring, assessment, access to dedicated forum,
management of the learning curriculum, are
introduced on the fly, depending of the situations
and needs. Transformations will be done at run-time
following rules and inferences about knowledge
modeling at design time.
There are two main situations for the
seller/learner:
The seller, or client, is outside the shop
counters: seller in the back office or storage
areas, client at home or another places;
The seller is in middle of his/her department,
alone or in front of potential client with the
possibility to use resources from the Smart
Space surrounding them.
We can precise here the second situation which
is richer in terms of interactions and importance of
contexts. In this situation, from the learning
perspective, the seller used the mobile information
system for both revising his/her knowledge about
products and selling characteristics or as a coach to
help him/her, in front of the potential client, in the
selling process especially in the selection/decision
phase. For this purpose he/she is equipped with a
new Ultra Mobile Personal Computer (UMPC) such
as the Q1 of Samsung, operating with WindowsXP
OS system and having a WiFi connection to the shop
network, and a Bluetooth to an earphone or
eventually to the mobile client’s phone (in order to
download information). From the earphone the seller
can receive calls from the call-centre of the shop or
enterprise, and computer generated voice message
with the help of the VoiceXML standard. Future
investigations will also support multimodal
interaction both using the digital pen or the touch
screen of the personal assistant and the microphone
included in the Bluetooth earphone. There are two
main variants:
One where the seller is also equipped with a
portable of a code-bar of RFID tag reader. This
enables the potential for real objects, such as the
product (DVD player for example), to play a
role into the interaction and the learning
process, because, they have their counterpart
into the virtual world;
The second where the seller can also
supplement his/her PTA with a larger LCD
screen, located in the department and connected
to the shop network. This is useful for the
sharing of information with the potential client
forming a Single Display Groupware (SDG)
with the PTA used as the coordination
supervisor. This is an example of SmartSpace
where some pervasive technologies augment the
ICEIS 2007 - International Conference on Enterprise Information Systems
218
potential of places with services and dedicated
apparatus shared by the user owners of mobile
devices.
This is a way to follow the vision about
Pervasive Computing as advocated by
(Satyanraynan, 2001).
4 THE DESIGN OF A FLEXIBLE
INFRASTRUCTURE FOR
P-LEARNING
4.1 Dynamic Composition of e-Services
for p-Learning with Multi-channel
Interactions
Following the movement already started in the field
of software engineering, with the Service-Oriented
Architecture (SOA) and in the business field, we can
show that SOA is also a good framework to reframe
the Learning and Content Management Systems
(LCMS). This is confirmed by proposals for the
future learning systems, seen as a collection of e-
Services, such as those of (Wilson et al., 2004) who
proposed a framework of e-Learning services for a
consensus at the international level. If we analyze
this proposal, we can see that the architecture, made
of numerous services, can be decomposed into three
layers from bottom to top: the underlying layer
concerns the data and their persistence across the
services relying on some databases. The
intermediate layer provides application services (for
education it is for example LearningFlow, e-
portfolio, authoring, etc.) and common services (AV
conferencing, role management, DRM…). The
upper layer not precisely described in (Wilson et al.,
2004) is called User Agents and is in charge of
managing the interaction between the user (through
his/her DUs) and a collection of services that
composed his/her applications.
This hypothesis about the importance of these
future open frameworks of learning services is one
of the main assumptions behind the design of our
infrastructure, with a focus on the upper layer. But,
because at this stage the whole framework of e-
Services is still under development, we have
decided, pragmatically, to “wrap” already available
services derived from existing e-Learning platforms
(LCMS existing in open source) providing learning
documents; a dedicated e-portfolio (http://elgg.org);
And from Wiki like one, in order to be compatible
with our specification of the e-Services and with the
Web services technology standards (I.e. the SOAP,
WSDL, UDDI trilogy). This is the left part of the
schema given in figure 2 about an overview of the
software architecture.
4.2 About the Modeling of Different
Contexts and Dynamic Adaptations
The e-Services proposed approach allows us to
concentrate our effort on the management of the
dynamic composition of e-Services directly in
interaction with the learners, that we have called
Interactive e-Services, IeS, in figure 2. We have
already shown (Chevrin et al., 2005a) that this
composition is not independent of the characteristics
of channel used, as usually think. For us, a channel
is the composition of a particular User Devices (i.e.
a personal assistant) with a particular
communication network, which proposes one or
more human communication modalities: textual,
direct manipulation through graphical interface,
voice….
Nevertheless, we have already done an important
research effort for the IeS composition and multi-
channel adaptation (Chevrin et al., 2005b). In order
to support more general mechanisms for the
adaptation of the flow of documents and interaction
elements to the different contexts, we have built a
meta-model of the context-aware process, following
for that purpose the proposals of (Henricksen, 2003).
Figure 2: Global view of our infrastructure.
A FLEXIBLE INFRASTRUCTURE FOR P-LEARNING: A FIRST APPLICATION IN THE FIELD OF
PROFESSIONAL TRAINING
219
This allows the description of the relationship
natures and the constraint associations, between the
different contexts and their effects on the main
elements used in this ontological model. This leads
to the design of our own adaptation metamodel that
we cannot present in detail here. In this metamodel
there are four main entities around the user/learner,
forming the core of the model, which are: the
Organization (responsible of the intention: training)
the Context, the Activity and the Channel. The
activity is linked to the learn-flow model that can be
expressed with the concepts of IMS-LD, our other
workflow metamodels.
In our collaboration with a Tunisian research
team, we have made a new step by considering the
concept of activity as part on the context. In fact, we
have proposed a new approach, which is the
bijective adaptation between contexts and user
activities (Malek et al., 2006). The next step of our
works is now to plug this “bijective adaptation
system” (symbolize by the bi-directional links in
figure 2) on our intermediation middleware for
supporting more pertinent ubiquitous interactions.
Up to now, it was the activities that were adapted to
the context. From now, from this point of view, the
context may be also adapted to the activities if it is
possible and pertinent.
We retain from the previous works on context-
aware computing that there is a need for better
models of contexts and models of adaptation of
services which are context-sensitive. At this stage
we reuse some results and technological solutions of
previous researches, for the parts of the context that
is not directly related to the learning situation. It is
the user profile, the technological context with an
emphasis on the channel context, and some parts of
the physical contexts, such as the detection of
position in front of a Single Display Groupware for
the seller of our proposed scenario in section 3.2.
Our effort is put mainly on the parts of the context
that is directly related to the learner activities.
Effectively the activity itself is also a context,
especially through the history of the past episodes of
learning and also the work activities. In the proposed
scenario this means that we must adapt the p-
Learning environment to the professional situation,
for example the nature of the present activities of the
sellers, and the presence or not of a potential buyer.
4.3 The Software Architecture for
Intermediation based on a
Multi-Agents System (MAS)
We present now our overall software architecture
based on a Multi-agents System. This is an evolution
of our previous infrastructure developed for research
in multi-channel and multimodal interaction in
mobile-Commerce (Chevrin et al., 2005b). This
architecture is dedicated to the management of the
intermediation between the IeS and the channels
used during an interaction. The central principle of
our solution is that the data, document, flows are
going from an abstract form (elaborated into the IeS)
to a concrete form, in the Application Server that
manages the different channels accesses, through the
chain of transformations, close to those given in
figure 1. The heart of our architecture, as given in
Figure 2, is implemented using the MAS technology.
The MAS role is to coordinate the use of the other
parts of the prototype, i.e. the context, the
organization policies, the intermediation between the
IeS and the channels used (synchronously or not)
during the interaction with the user, etc. The
adaptation of the IeS is done through the
intermediation middleware already presented in
details in (Chevrin et al., 2005b).
The interest, of using a MAS as a substrate for
the design and implementation of our infrastructure
for p-Learning, is of course due to the already well
known flexibility, and potential to support evolving
systems, of MAS as dedicated middleware for
pervasive computing. But less known the multi-
agents paradigm is also useful for some complex and
dynamic mechanisms required in multi-channel and
multimodal interactions with the end users, the
learners. The two main mechanisms are:
The fission mechanism: where the abstract
document can be split into contextual sub-
elements that are adapted for, and routed to, the
right channel. This form the multimedia output
in a multi-device context (more in section 5);
The fusion mechanism: where two input
channels (i.e. voice and graphical selection)
used during an interaction with the user, a
session, are combined, in order to form a
request: understandable by the middleware
agents and directed to the right service. This is
the multimodal mode of interaction.
For the MAS we have used a technological
platform that is relatively mature now, JADE
(http://jade.tilab.com/), in operation in several
projects on mobile communications supported by
telecommunication providers, and that offers the
possibility to deploy software agent even inside the
User Device, due to its implementation in JAVA,
and its respects of technological standards for
interoperability.
ICEIS 2007 - International Conference on Enterprise Information Systems
220
5 EXAMPLES OF ADAPTATION
FOR THE PTA SCENARIO
In the proposed scenario, the seller as a learner can
do three types of activities, which are
contextualized:
Access to the learning contents by navigation
guided by a network of concepts relative to the
selling domain (ontological model and adaptive
hypermedia);
Access to the learning contents through the
product datasheets or frequently ask questions;
After a dialogue with the client, selection of a
similar case already solved (for example
problem of a home studio configuration), which
help her/him to give a better advice to this client
(Case-Based Reasoning);
In one of the phase of the scenario, the seller
asks the client to follow him/her to a specific
apparatus placed inside the store department in order
to augment the potential of his/her PTA. In this
pervasive learning situation, the fusion and fission
mechanisms are used to support the collaboration of
the user device, his/her PTA, with the SDG (see
figure 3 for the setting), that requires a fission
operation, or even with the mobile communication
system owns by the potential client (i.e.
downloading of commercial information). The data
and documents provided by the IeS of the left part of
the architecture as described into figure 2, must be
dynamically split into different streams depending of
the contexts, for example the user preference and the
present characteristics of the two or three channels
in action. Same the use of a voice channel in input
jointly with user action on his/her tactile screen of
the UMPC required a fusion operation in input in
order to analyze the request in accordance, also, with
the present contexts of interaction. In this case, we
will have several particularities such as using multi-
devices and “physical objects” of the environment
such as the shared display or RFID tags on products.
The Figure 3 shows an example of SDG
arrangement. This SDG is composed of four distinct
channels: two are visual, one small and private
viewed by the seller, one large and public viewable
by several persons; the voice channel directed to the
seller, and private, thanks to the Bluetooth auricle;
and possibly the data channel of the client mobile
system if connected, for example by Bluetooth link,
to download commercial documents and guidelines.
The heart of this scenario implementation is the
dynamical fission of the different information on the
different channels. Projects like (Han et al., 2000)
deal also with this subject.
The Figure 2 shows the different software agents
implemented to manage the fission mechanisms. The
Session Agent keeps a state of the different channels
used, and in this way, the data fission. For the
moment, only one software agent (Proxy Agent),
manages the routing to the dedicated channels.
Figure 3: Example of SDG arrangement in the PTA
scenario.
6 CONCLUSION
The application of advanced technologies and
models introduces the possibility to dynamically
adapt the contents and the services composition to a
particular learning and learner situation, in
accordance with various contexts, including his/her
past experiences as learner and the present situation
as worker, and the impact of his/her mobility on
location, proximal resources… The new step, called
pervasive learning environment, is to realize the
potential of the pervasive computing vision
(Satyanraynan, 2001), which proposes the extension
of mobile computing with the potential of
Smartspace, where the user interaction and
experience can be enriched with concrete objects,
having their computerized counterparts, and by e-
Services embedded into the surrounding (for
example the office, the shop…) and supported by an
invisible and very proactive, attentive, infrastructure.
However the potential of designing and using
pervasive learning environments is hampered by two
interrelated problems: First to find real needs and
significant situations for this proposition; And
second by the lacks of open infrastructures, and
equivalent of the past LMS, adapted to context-
aware dynamic adaptations of contents and services,
and the multi-channel and multimodal nature of the
user interactions that are inherent to the use of small,
limited, embodied, mobile devices.
We think that professional world with the
demand for continuous training and learning on
A FLEXIBLE INFRASTRUCTURE FOR P-LEARNING: A FIRST APPLICATION IN THE FIELD OF
PROFESSIONAL TRAINING
221
demand of workers in real situations (professional
activities and contextual physical settings) is a good
candidate for an application of the potential of p-
Learning. For that purpose we are developing a first
solution, called the Personal Training Assistant,
supporting the counseling and selling of products
that are complex to master and in continuous
evolution, the sector of TV HD in our case. This is a
new learning scenario that has a great interest for the
retail industry and its evolution to e-Retail.
Our contribution, to the second problem solving,
is to propose an infrastructure sufficiently open and
flexible to support a wide range of p-Learning
settings. However the difficult problem is about the
reuse of the previous different contexts models and
the standardization of the information they can
provided. In p-Learning we think that this must be
done pragmatically by reusing Context Provider
Services when they are available, for example as
results of more general researches in pervasive
computing, or even proposal from international
consortium such as Liberty Alliance for the
management of the digital identity of the learner
including his/her profiles as users of e-Services.
As researchers in the field of Technology
Enhanced Learning, we must concentrate our
collective effort on the definition and modeling of
the part of the contexts that are particular to the
learning activities and processes. This elicits the
weakness of currents standards for e-Learning, such
as SCORM for the learning objects or IMS-LD for
the pedagogical scenario and learning activities,
because they offer no possibility to specify and
support dynamic adaptations that are context-aware.
ACKNOWLEDGEMENTS
The present research work has been supported by the
“Ministère de l'Education Nationale, de la Recherche
et de la Technologie», the «Région Nord Pas-de-
Calais» and the FEDER (Fonds Européen de
Développement Régional) during the projects
MIAOU and EUCUE. The authors gratefully
acknowledge the support of these institutions. This
has been also supported partially by the p-LearNet
project funded by the Agence National de la
Recherche.
REFERENCES
Attewell, J., 2005. Mobile Technologies and Learning: a
technology update and m-learning project summary.
Document from the “Learning and Skills Development
Agency”, UK, (www.lsda.org.uk).
Bomsdorf, B., 2005. Adaptation of Learning Spaces:
Supporting Ubiquitous Learning in Higher Distance
Education, Mobile Computing and Ambient
Intelligence: The Challenge of Multimedia, Dagstuhl
Seminar Proceedings.
Chevrin, V., Derycke, A., Rouillard, J., 2005b, Some
issues for the Modelling of Interactive E-Services
from the Customer Multi-Channel Interaction
Perspectives. Proceeding of the eee05 international
IEEE conference, pp 256-259.
Chevrin, V., Rouillard, J., Derycke, A., 2005a. Multi-
channel and multi-modal interactions in E-marketing:
Toward a generic architecture for integration and
experimentation, HCI International 2005, Lawrence
Erlbaum Associates editor, 10 pages.
Han, R., Perret, V., Naghshineh, M., 2000. WebSplitter: a
unified XML framework for multi-device
collaborative Web browsing, Proceedings of the ACM
conference on CSCW, December 2000, Philadelphia,
Pennsylvania, United States, p.221-230.
Henricksen, K., 2003. A framework for context-aware
pervasive computing applications. PhD thesis. School
of Information Technology and Electrical Engineering.
The University of Queensland.
Jones, V., Jo, J., 2004. Ubiquitous Learning Environment:
an Adaptive Teching System using Ubiquitous
Technology. In R Atkinson, Mcbeath, Joans-Dwyer
(eds) Beyond the comfort zone: proceeding of the
ASCILITE conference, Perth, Australia, 5-8 December
2004, pp 468-474.
Keil-Slawick, R., Hampel, T., Ebman, B., 2005. Re-
Conceptualising Learning Environments: A
framework for Pervasive eLearning. Proceedings of
3
rd
Int’l conf. on Pervasive Computing and
Communication (PERCOM), IEEE press, 6 p.
Malek, J., Laroussi, M., Derycke, A., 2006. An Adaptor
for Bijective Adaptation between Context and
Activity in a Mobile and Collaborative Learning,
IADIS International Conference Mobile Learning,
Dublin, Ireland, 14–16 July 2006.
Satyanaryanan, M., Pervasive Computing: Vision and
Challenges. In IEEE Personal Communications, IEEE
press, August 2001, pp 10-17.
Sharples, M., (2003). Disruptive devices: mobile
technology for conversational learning. International
Journal of Continuing Engineering Education and
Lifelong Learning, 12(5/6): 504-520.
Syvänen, A., Beale, R., Sharples, M., Ahonen, M.,
Lonsdale, P., 2005. Supporting Pervasive Learning
Environments: Adaptability and Context-Awareness in
Mobile-Learning. IEEE WMTE 2005 conference
proceedings, pp 251-253.
Wilson, S., Blinco, K., Rehak, D., 2004. Service-Oriented
Frameworks: Modeling the infrastructure for the next
generation of e-Learning Systems. Paper for the Alt-I-
Lab 2004 conference.
Woodill, G., Emerging E-Learning Technologies: Tools
for Developing Innovative Online Training. 2006.
Report, 344 pages. (www.brandon-hall.com).
ICEIS 2007 - International Conference on Enterprise Information Systems
222