IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
18
SMARTNotes : Semantic annotation system for a collaborative
learning
Driss BOUZIDI 1, Rachid ELOUAHABI2, Noreddine ABGHOUR 3
1
Department of mathematics and computer Science, Hassan II Ain Chock University,
Faculty of Sciences, Casablanca, Morocco
2
3
Department of mathematics and computer Science, Moulay Ismail University,
Faculty of Sciences, Meknes, Morocco
Department of mathematics and computer Science, Hassan II Ain Chock University,
Faculty of Sciences, Casablanca, Morocco
Abstract
Creating situations that promote collaboration between learners
and/or their tutor is generally based on traditional
communication tools. These are an important means to exchange
ideas among learners, validate and consolidate their learning.
However, we note that the large volume of messages exchanged
between students and the tutor generates unwanted noise that can
cause problems of disorientation for the majority of learners and
cognitive overload.
Through a solution based on the use of annotations as a support
for collaboration, we sought, first, to stimulate interaction and
facilitate collaborative activities between learners and their tutor
in an activity of understanding of a distance-learning course.
Secondly, we have proposed to create connections, through
automatic annotations, linking parts of the course to the most
relevant messages posted in a discussion forum. The degree of
relevance of these messages is based on a customized
classification according to the profile of the learner and the
objective of the course, as well as the integration of a semantic
search done by applying a thesaurus to the LSA method.
Keywords: Annotation systems, collaborative learning,
discussion forums, messages classification, Semantic Web,
Latent Semantic Analysis, LSA.
1. Work context and problematic
On the basis of the logic of classical training, where the
learner who wishes to understand the concepts of a course,
taking full advantage of the discussions that take place in
the classroom to dispel ambiguities and/or more detail on
these concepts, the e-learning solutions have encased the
training content of technological tools of communication.
These systems gather in one place all the necessary tools
to learners and tutors to follow learning activities. These
tools enable communication (e-mail, forums, chat, etc...),
share resources and files (shared bookmarks, virtual
libraries, etc.), and even offer distance courses (the case of
videoconference session, etc.)[1].
These tools are an important means to conceal the feeling
of loneliness within the learner, and encourage interaction
with the tutor and peers, this allows him to have/provide
support of/to other learners [2].
However, this multiplicity of tools for communication and
information sharing, was not without its drawbacks, the
fact that distance training often takes place in
asynchronous session (the learners are not obliged to
follow courses at the same time), usually when the student
logs in to the online learning platform, he is faced with a
large number of messages that emerged in an exponential
and uncontrolled way in the communication spaces,
during its disconnection [3].
Generally we see that the important number of messages
exchanged between students and tutor to generate
unwanted noise that is proportional to the number of
participants [4]. Read all that was exchanged in these
different sources of information becomes a difficult and
harmful task. It can, therefore, lead to problems of
disorientation for the majority of learners and induce
cognitive overload.
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
Also searching for information in these environments is a
difficult task that requires more patience, the learner finds
it difficult to distinguish the types of messages and know
their level of educational importance (personal exchanges,
administrative exchanges, discussions on the content,
group exchanges, etc).
In addition, a need for support is often felt during the
progress of the learner in understanding the course[5], the
fact that the messages exchanged between learners
relating to parts of the course are scattered in various
communication spaces (mail, forum discussion, chat, etc.)
and so set outside course document, establishing a link
between the messages exchanged in these different
knowledge bases and content of the source course of these
exchanges, generally goes through external references to
the course document. The learner finds it difficult to
establish the connections between the messages produced
mainly in the discussion forums and electronic mail, and
the parts of its course. This tends to make the learning
process much more cumbersome, inducing cognitive
overload for the learner.
The learner finds himself lost in this rich mine of
information but unfortunately not easy to handle, pushing
him not to use it properly as a source to supplement and
enrich his learning process.
In this article we present the specifications of our
annotation tool called SMARTNotes to support the learner
in learning activities, and we look at the problem of
linking the digital course materials and exchanges
produced through communication tools, particularly
discussion forums. We propose an approach to
automatically generate annotations on the course
document leading to messages produced through the
communication tools. This will allow on the one hand to
remain open on the choice of communication tool within
the constraints of the learning activities and also to avoid
the learner the effort to link between the course document
and exchanges of posted messages in the bases of these
communication tools.
2. Specification of SMARTNotes annotation
system
Because communication tools, find their relevance only if
they are used in a context providing the best possible
interactions with respect to exchanges around the course
content, we propose a solution that incorporates the
messages produced by learners and the tutor in this latter.
It was inspired by the practice of annotation of paper
documents, which is a practice frequently used by readers
to write personal comments on its margins [6] [7]. This
activity allows annotators to express their views, to build a
stable and written memory of the passages that they
annotate [8].
The annotation on the Web offers more to the reader the
opportunity to share his notes. Thus, the person consulting
19
an annotated document can acquire annotations attached
to this document. This allows asking questions, giving
advice and discussing problem solving in group, by using
the annotations. The learner can have a complete and an
enriched overview of his course document containing its
messages (advice, questions, answers, references, etc...),
and those of others attached to their corresponding parts in
the document. This will save the effort to the learner to
make associations between his notes (and those of his
colleagues), and parts of the course sources of these
interactions.
It is in this direction that we have proposed a collaborative
annotation system, we have called SMARTNotes. This
system aims to provide learners (and tutor) with support
tools for fostering collaboration through the course
document. The learner can interact, validate and enrich his
course by the notes generated during the collaboration
with his tutor and his peers.
2.1 Definition and objective of the annotations
The annotation is an action performed by setting a mark
on an object. It is a sign of mental state that the reader gets
about the annotated item [7]. In our context, the latter
representing the target to which the annotation is linked,
can be a collection of documents, a document, or any part
thereof (paragraph, sentence, word, image ...), or even
another annotation.
Studies on annotations [5][9][10], have shown that
annotations made in a shared document can be used either
by the annotator himself for personal use (support for
active reading, customization, argumentation, etc.), or a
consultant of the annotated object, be it a human
(ownership, guidance and counseling, discussion and
collaboration on the document, etc.) or a machine
(automatic generation of abstract indexing for extensive
research, tracking interactions, etc.).
2.2 Collaborative annotation systems
Several annotation systems have been developed, some
addressing common issues and others specific to them.
There are two types of field of use of such systems, either
to index web resources to facilitate research, or to
facilitate communication in an activity of understanding a
document or the completion of a joint work involving
several actors. Among these systems, we can mention,
Annotea which is an experimental project of consortium
W3C aiming to develop an environment of shared and
collaborative annotation [11]. By using open standards of
W3C, in particular RDF (Resource Description
Framework), it arises as an objective to promote
interoperability between applications that exchange
annotations in the form of metadata. CoNote is also a
collaborative annotation system developed at Cornell
University [12]; it focuses on rights of access to
annotations for a group of people who share a document.
The annotations in the Yawas system are dedicated to the
automatic and customized classification of annotated
documentation. It shows that the classifications obtained
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
through annotations are more accurate than those obtained
using the full document [13].
2.3 SMARTNotes collaborative annotation tool
In our SMARTNotes system, the annotator (especially a
learner) interacts with the document he consults ; he is
actively involved in enrichment. He is no longer seen as a
simple passive reader, but he becomes more of a reader / a
writer of the document; he completes it by his own
understanding by annotating it as and when he progresses
in reading. These notes serve as a means of locating
relevant information and clarification in the form of
comments.
2.3.1
The annotation representation model
One of the basic elements in the establishment of an
annotation system is how to organize information about
the object in the annotated document to be manipulated
and implemented without any difficulty.
We were inspired by the model proposed by the W3C
consortium which is an annotation as a set of metadata
(attributes of annotation) and the body of the annotation
(the content of the annotation). Properties induced by this
model are, firstly, the opening notes to share with other
systems that meet this standard and also the possibility of
the scalability of this model to support new annotation
types.
Metadata are defined as an RDF schema. This latter
represents a set of specifications aiming to standardize the
modeling of annotations to ensure interoperability
between applications exchanging metadata.
Fig. 1 RDF model of an annotation proposed by W3C
The annotation shown in the figure above was created on
27-02-2011 by the AnnotatorX, and subsequently
amended on 15-03-2011. It is associated to the context
located in the Xdoc.html document and its content is put
into the document DocAnnotation.html.
2.3.2
content of any posted messages if he wants to recognize
the learners with learning difficulties (ask more questions)
from those who are not (offer more responses).
We feel it is extremely interesting to type annotations in
SMARTNotes we defined taxonomy of acts of dialogues
adapted to the type of interactions that the annotator can
do. It is proposed to extend the RDF model to also support
the type of the annotation. We categorized all the
annotations according to their purpose and their
semantics. Each category is associated with specific types
expressing the subject of the action of the annotator.
When an annotator (learner/tutor) decides to create an
annotation, the system asks him via a semi-structured type
of annotation to be set. This technique did seem to us a
little heavy for the annotator, but it is most relevant in the
field of communication. On the one hand, the fact of
assigning a type to annotation leads the annotator to ask
about the specific purpose of this [14]. On the other hand,
this semantization will enormously facilitate the automatic
analysis of the behavior of the annotator in his group [15].
Category
TYPE
DESCRIPTION
Important
Allows emphasizing the
selected part to give more
value compared to the other
parts of the document. The
annotator can explain, with a
comment, the cause for which
he judges that a passage
Highlighting
Highlighting is important.
Useless
Allows to cross out the
selected part to indicate that it
is useless. The annotator can
explain, with a comment, the
cause for which he judges
that a crossed out passage is
useless.
Advice
The annotator can propose
advice in the form of
annotation to direct another
annotator. For example, a
tutor (even a learner) can give
Consign
advice to another learner who
is facing some learning
difficulties.
Explanation Allows to add an explanation
Definition of semantic annotation
The formats used to annotate a document differ from one
annotator to another; each makes his annotations
according to his own semantics. This can lead to an
overload in reading annotations; a learner cannot know the
subject of an annotation (question, answer reference, etc.)
unless he reads its contents. Also, the tutor has to read the
20
Conversation
Question
Allows to ask question
Answer
Allows to answer a question
Discussion
Allows to open a discussion
forum which can be carried
out between two or several
annotators on an annotated
object. An annotation of the
discussion type is a made up
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
Reference
Indication
Alert
of annotation which can
include an annotation of the
question type, answer, advice,
explanation.
Represents a reference to
another resource that can be
an internal passage in the
document, an annotation on
the document or to an
external document. This
allows the reader to refer to a
reading in relation to the
passage annotated.
To do this, we proposed an approach based on two steps:
the first allows a semantic classification of posted
messages through the communication tools. This
classification is based on: (1) the creation of a thesaurus
based on the interests of the learner and the objectives of
the course, (2) adapting the LSA method (Latent Semantic
Analysis) to group the posted messages with thematics
that are semantically close.
The second step of our approach is to automatically
generate annotations linking parts of the course document
to the messages exchanged in discussion forums in
connection with these parties.
allows the annotator to plan
future temporal actions
(annotation performed by the
system according to a given
date) and non-temporal (type
of annotations to read).
4. Semantic classification of posted messages
in a discussion form
Table 1. The categories and types of annotations in SMARTNotes
2.3.3
21
Annotation based discussion forums
We have proposed in the first version of SMARTNotes
collaboration support between users through chat rooms
based annotation. The annotator can post a message as a
comment, question and / or response by annotating
annotation. The conversational type of annotation is then
presented as a discussion thread. The annotations are
attached together and presented as a tree structure similar
to the classic discussion forums. The advantage of this
approach is that the context of the discussion-forum
annotation based is defined by default by the content of
the annotated part. This allows for even easier and direct
exchange on the parts of course document.
3. Automatic semantic annotation of course
documents
As we have previously reported, the communication tools
have an important place in an e-learning platform; they
have a very important role in stimulating interaction
between the learners and their tutor. Each tool has its
place in the learning activities according to many factors:
objectives sought, characteristics of the target audience,
time and technical constraints. For these reasons, we
found it beneficial to create links between our
SMARTNotes system and other communication tools. We
thought it better to integrate on our system the ability to
generate automatic annotations filed on the course
document leading to the messages produced through these
tools. This will allow on the one hand to remain open on
the choice of communication tool according to the
constraints of the learning activities and also to enable the
learner avoid efforts to make the connections between the
course document and exchanges of posted messages in
the bases of these communication tools.
To facilitate the search of exchanged messages, the
majority of communication tools use a classification based
on keywords chosen by the learner; the results returned
are generally independent of the conceptual intentions and
areas of interests of the latter; this makes them in most
cases ineffective and unintelligible. These problems are
increasing on the one hand with the volume and variety of
messages exchanged in discussion forums and also
because of the synonymy and/or polysemy problems.
We propose to include in SMARTNotes a semantic search
process delivering messages according to the most
appropriate interests of the learner and the objectives of
the training undertaken. This research does not require
direct involvement of the learner. All relevant information
to the search query are implicitly acquired from the profile
and training objectives set by the author of the course.
As mentioned earlier, the important volume of messages
exchanged through the communication tools often
generate unwanted noise, making their reading a difficult
and non practical operation. The purpose of our work is to
better exploit its messages for an instant support to the
learner in his activities of construction of knowledge
through the course document.
4.1 Support for the learner
classification of messages
profile
in
the
The user profile is the subject of attention in several areas
in particular that of education. It is represented by a set of
educational and general information on the learner that are
useful to establish an adaptive learning. The specifications
of standard models have proposed structuring the
information into categories: identification, access,
relationships with others, etc.. In this article, we are
interested in the IMS-LIP (Instructional Management
Systems Global Learning Consortium for Learner
Information package) standard, which offers a rich profile
model used in most current systems of learning, and
having the category "GOAL" describing personal learning
objectives and aspirations of the learner.
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
We then propose to exploit this category to achieve an
automatic classification of messages exchanged between
learners, which will be adapted to the learner profile.
For a better classification of messages, we also add to the
information in this category keywords associated with
learning objects. These keywords are defined by the
author of the course during the creation of the latter. We
have adopted the model proposed by LOM (Learning
Object Metadata) is a de facto standard and widely used.
LOM provides the element of "Keyword" in the category
"GENERAL" dedicated to store the keywords related to a
learning object.
22
To build our corpus, we did a search on each message to
select the terms which include more information. As a
result, we have created the semantic relationships
(hierarchical, equivalence and association) between these
terms.
All keywords deducted of "GOAL" of the learner profile
and the "Keyword" element of learning objects (the course
document), is the query we will use to search for posted
messages in the discussion forum.
At the end of this process, we obtain a global organization
of all the terms of our corpus according to semantic
relationships that will generate the basic thesaurus for our
research.
4.2.2
Fig. 2 Process of automatic generation of annotations
4.2 Semantic classification of messages based on
LSA
Instead of searching the messages based solely on the
terms set by the learner profile and keywords associated
with learning objects, resulting in most cases in restrictive
findings because of synonymy and polysemy problems,
we propose (a) to extend this research to the terms
semantically related to them by referring to a thesaurus.
(b) classify messages by using the LSA method that will
be completed by the measure of similarity of messages
based on the application message [16].
4.2.1
Process of creating the thesaurus
The thesaurus is as an instrument of control and
structuring of the vocabulary; it contributes to the
consistency of indexing and facilitates the search for
information to modulate the rate of recall and precision in
the identification [17].
We then created a thesaurus in the area of information
technology from the whole corpus of messages in
different topics (e.g., system, language, technology, etc.).
Classification of messages by using LSA method
The Thesaurus, designed in the first stage, will be applied
in the construction phase of the lexical table (Terms /
messages) that will precede the application of the LSA
method (Latent Semantic Analysis).
The LSA is a method to determine similarities between
documents; a document can be a text, a paragraph or even
a sentence or a word [18]. For this, each message is
represented by a vector m= (d1, d2, d3, …, dn) called
lexical profile in which the jth component dj represents
the weight (or importance), the message m, the indexing
term tj associated to the ith dimension of the vector space.
Then we proceed to the normalization of this matrix [19].
For successful application of the thesaurus during the
construction phase of the lexical table we have adopted an
approach which is to include more keywords mentioned
by the user, the specific terms associated with them via
the thesaurus and common to them while avoiding
repetition.
T1 and T2 two terms mentioned by the user. T11, T12
T13, are terms specific to T1 and T21, T12, T11, T22specific to T2. We note that T11, T12 are common to T1
and T2.
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
23
introduced by the user without being given in the latter
[19].
5. The SMARTNotes architecture system
5.1 The annotation system architecture
The generated lexical table as follows :
Messages
T1
T11
T12
T2
Occurrences of terms of each
message
Table. 2 Lexical table includes all the terms in semantic relationships
with keywords
For all the messages, we take into consideration all those
coming from our basic message and the query keywords
that will be considered also as a message. To classify the
messages we use the LSA method that will be completed
by the similarity measure of basic messages with the
message request. To do this, we chose the cosine which is
a simple measure in terms of computation and accurate in
terms of results compared, for example, to the Euclidean
distance and the scalar product [20].
Following the first approach, the overall architecture of
our system can be summarized as per the following
diagram:
Fig. 4 General architecture of semantic classifier module
The application of this approach shows that the research
carried out by applying a thesaurus to the LSA gives a
more relevant result than that obtained from the
application of the LSA only. The results obtained return
messages whose context is one of the themes of the terms
The standard architecture of a system of annotation is
primarily based on the notion of intermediary that
provides the interface between the client and the Web
server [13], [23]. It is responsible for processing any
transaction on the annotations between client/server and
consists of four complementary modules to ensure the
execution of the annotation:
The Interceptor is solicited at each request sent by the
client browser. If the application requests the loading
of a page, the interceptor sends it to the Web server,
get the result, then needles is to the Composer module.
The composer is responsible for include the
annotations extracted from the annotations base (if
any) in the requested page and returns the result to the
Interceptor module. The latter returns the annotated
page to the client so that it can finally be loaded by the
browser.
The Annotation Management module is called by
the interceptor to update the database of annotations
on requests from the client browser to do so.
The User Management module provides the
management of users and access rights to the
annotations.
We have distinguished three types of implementation of
the architecture of an annotation system according to the
position where the intermediary is installed:
The first architecture proposes the intermediary to the
Web server. The proposed annotation features are
limited to web pages published on it
The second architecture places the intermediary on a
particular server set independently of the client and
Web servers. This proxy server follows the standard
pattern of an annotation system; it acts as an interface
between the client and Web servers and manages
pages with annotations on its base [24]. Among the
weaknesses of this architecture is that we mainly
found the response time slow. All tasks are performed
by the proxy server, the search for the requested page,
the extraction of annotations associated with it from
the base, the integration of these annotations on the
page, and return the response to the browser applicant,
make of this server a major choke point.
In the third architecture the intermediary is put near
the client, as a plug-in or external application (eg an
applet). The aim is to enrich the browser functions to
manipulate the annotations of a web page. This
architecture has more advantages than the previous
two: ability to annotate web documents stored locally,
more flexibility and control of interactions of the
annotator and avoiding the bottleneck at the server.
However, opposed to this, it remains dependent on the
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
type of browser used and the possibilities of sharing
annotations and collaboration on web documents are
painfully managed.
5.2 Main components of SMARTNotes system
In SMARTNotes, we propose an architecture based on an
intermediary divided between the client and server. The
first part is set at the browser as an extension, allowing the
user to annotate a document, when loaded by the browser,
without using the annotation server. The interceptor
module provides communication between the client and
the server; you can either retrieve the annotations
associated with the document loaded from the server or to
transmit the updates made. The composer module is also
placed on the browser; it is based particularly on the DOM
and XPointer technologies to include annotations in the
web page being loaded.
The second part of the intermediary server is placed on the
annotation to intercept client requests and manage the
annotations present in the base. The semantic
classification module of the messages posted in the
discussion forum enables to deliver the most appropriate
messages according to the interests of the learner and the
objectives of the training undertaken.
24
server; this will also avoid the problem of the
bottleneck at the server level annotation.
Sharing of annotations and collaboration on the
document in a simple way: they are based on export
operations and synchronization between the server and
annotation client. Each time the server intercepts an
update of a page, it sends a message to the client
connected to the server and logging on the same page.
The automatic refresh of the page is done on the client
terminal only after confirmation of the user warned.
Response time very small: the set up of the composer
and the manager of annotations at the level of client
allows on the one hand to manipulate the annotations
of pages with very low response time. Secondly,
reduce the frequency of use of the annotation server as
backup and search annotation on its base. And
therefore also have long response times on the server
annotation.
5.3 Processus of Integration of annotations in the
course document
In SMARTNotes, communication between the client and
the server is based on an exchange annotation XML. We
have defined a database link associating the content of the
annotation, which is stored in the annotations or
discussion forum, and the corresponding part of the
course. This solution enables the reader to view the
document associated with annotated annotations in a
transparent manner and unchanged the source document.
Fig.6 Base of links joining the basis of exchange information to the
course parts
This architecture has several advantages, namely:
Respect of confidentiality: providing the ability to save
annotations locally with the client.
The off-line annotation of documents: the integration
of the intermediate level clients can annotate a
document (local or remote after the download on the
browser) without having to connect to the server
annotation. The annotator can make a backup of
annotations locally or export batch annotation on the
The connecting links between the base of exchange
information and to the document to be annotated are put
into a database of links regardless of the course document
which solves the problem of copyright of the course.
Thanks to the XLink technology which offers new
mechanisms making hypermedia documents more
flexible, we defined this XML document called "basic
links" without text links that are totally separated from the
source document.
6. Conclusion and perspectives
In this work we have presented barriers that a learner may
face in the misuse of traditional communication tools in
an activity of understanding of a distance-course. In
particular, the large volume of messages posted in these
IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011
ISSN (Online): 1694-0814
www.IJCSI.org
areas can lead to disorientation for the majority of learners
and cognitive overload. We have proposed to put the
hypermedia course documents in a broader dimension to
centralize much more interaction and exchange of ideas
for successful learning. Through a solution based on the
use of annotations as a medium for collaboration, we
sought, first, to stimulate interaction and facilitate
collaborative activities between learners and their tutor in
an activity of understanding of a course. Secondly, we
have proposed to create connections, through automatic
annotations linking parts of the course to the most relevant
messages posted in a discussion forum. The level of
relevance of these messages is based on a customized
classification according to the profile of the learner and
the objective of the course as well as the integration of a
semantic search done by applying a thesaurus to the LSA
method.
The design of the functional architecture of our annotation
tool has been defined; it meets a number of requirements
such as independence from the platform hardware /
software, interoperability, scalability to support other
types of annotations. Also, in order to verify the feasibility
of our proposal based mainly on XML technology, unit
tests validation were successfully completed. The
application of our approach to semantic classification on a
corpus of messages posted through a discussion forum of
the MOODLE platform has shown relevant results with
the LSA method.
We are now in a period of experimentation with our
system and we plan to include a recommendation system
that draws in the documents annotated by learners and
tutors, and automatically offers educational resources for
active learners without having to explicitly request their
feedback.
References
[1] N. ABGHOUR, "Schéma d’autorisation pour les applications
réparties sur Internet", Thèse de l’Institut National Polytechnique,
Toulouse, Rapport LAAS, 2004
[2] Henri, F. & Lundgren-Cayrol, K « Apprentissage collaboratif à
distance, téléconférence et télédiscussion.» Rapport interne
n°3.Montréal, Canada; Centre de recherche LICEF, 1996.
[3] D.Bouzidi , R.Ajhoun, A.Benkiran, "The use of the annotations in a
collaborative activity of reading an online hypermedia course
document". IEEE 1-4244-0406-1, 2006
[4] Marc Walckiers and Thomas De Praetere « L’apprentissage
collaboratif en ligne, huit avantages qui en font un must ». Vol.2
Distances et savoirs 2004, ISSN 1765-0887.
[5] R.Elouahbi A.Nassir and S.Benhlima, "Design and implementation
of adaptive sequencing based on graphs", 11th International
Educational Technology Conference, Turkey, 2011, p 233-238.
[6] C. Marshall, "Annotation: from paper books to the digital library ", In
Proceedings of DL ’97, 1997, pp.131-140.
[7] M.Veron « Modélisation de la composante annotative dans les
documents électroniques » Rapport de stage du DEA Représentation
des Connaissances et Formalisation du Raisonnement, UPS-IRIT,
Toulouse, Juin 1997.
[8] D.Bouzidi, « Collaboration dans les cours hypermédias du système
de télé-enseignement SMART-learning », Thèse de Doctorat,
Spécialité Informatique, Ecole Mohammadia d’Ingénieurs (EMI),
Université Mohammed V-Agdal, Rabat, 2004.
[9] J.J Cadiz, A.Gupta, J. Grudin "Using Web Annotations for
Asynchronous Collaboration Around Documents", Conference on
25
Computer Supported Cooperative Work (CSCW2000), Philadelphie,
Pennsylvanie, USA, 2000.
[10] L. Denoue, « De la création à la capitalisation des annotations dans
un espace personnel d’informations ». Thèse de doctorat (PhD),
Université de Savoie, Octobre 2000.
[11] M.R Koivunen, "Annotea : Applying Semantic Web Technologies
to Annotations", in "Semantic Web Kick-Off en Finlande, Vision,
Technology, Research and Applications", HIT Publication, Hyvonen
Eero (ed), p.213,226, 2002
[12] J.Davis and D.Huttenlocher, "Shared Annotation for Cooperative
Learning". Proceedings of the Conference on Computer Supported
Cooperative Learning, 1995
[13] L.Denoue & L.Vignollet, "Yawas un outil d’annotation pour les
navigateurs du Web", IHM’99Montpellier, France, 1999.
[14] G.Sébastien, "Apprentissage collectif à distance. SPLACH : un
environnement informatique support d’une pédagogie de projet."
Thèse de Doctorat, Spécialité Informatique, Université du Maine,
2001
[15] D.BOUZIDI, R.AJHOUN, A.BENKIRAN, "Les annotations pour
un
apprentissage
collaboratif",
JOCAIR’2006,
colloque
International, France 2006 pp.372-382.
[16] S.LGARCH, D.BOUZIDI, S.BENNANI, "Une approche
sémantique de classification de messages d’un forum de discussion
basée sur la méthode LSA", WNGN Workshop on Next Generation
Network, pages : 3-8, Maroc 2008, pages 3-8.
[17] L.Saadani et S.Bertrand-Gastaldy, " Cartes conceptuelles et
thésaurus : essai de comparaison entre deux modèles de
représentation issus de différentes traditions disciplinaires ",2000
[18] Y.Bestgen , « Analyse sémantique latente et segmentation
automatique des textes » , 2004
[19] G. Salton, « Autoamtic Text Processing – The Transformation
Analysis and Retrieval in Full Text Information Systems», 1989
[20] H.Zargayouma « Indexation sémantique de documents XML »
Thèse de doctorat, 2005
[21] S. LGARCH, "Une approche sémantique de classification de
messages d’un forum de discussion basée sur la méthode LSA",
Rapport de DESA, Mohammadia Engineering School, Mohamed V–
Agdal University, Rabat, Morocco, 2008.
[22] V.Vasudevan and M. Palmer, « On web annotations : Promises and
pitfalls of current web infrastructure». Actes de 32nd Hawaii
International Conference on System Sciences, Maui, Hawaii, 1999
[23] I.Ovsiannikov, M. Arbib, et T. McNeill, « Annotation technology».
Int. J. Human-Computer Studies, 50 :329–362
[24] E.Desmontils, C.Jacquin & L.Simon, « Vers un système
d'annotation distribué », Rapport de recherche n°03-01, Nantes,
Institut de Recherche en Informatique Université de Nantes, 2003,
16 p.
Driss BOUZIDI is an Assistance Professor in Computer Science at Hassan
II Ain Chock University, Faculty of Sciences Casablanca, Morocco. He
received a Ph.D. degree in computer engineering from Mohammed V
University in 2004 on "Collaboration in hypermedia courses of e-learning
system SMART-Learning". Dr. Bouzidi's research interests are in
Distributed multimedia Applications, Collaboration systems, and e-learning.
He was vice-chair of the international conference NGNS'09 (Next
Generation Network and Services) and the treasurer of the two research
associations e-NGN and APRIMT.
Rachid ELOUAHBI received his doctorate degree in Computer Science
from Mohammadia School of Engineering Morocco in 2005. He is currently
an Assistant Professor at the University of Moulay Ismail, Meknès. His
research focused on the area of adaptive learning systems, course
sequencing and learning technologies. He works on the project Graduate's
Insertion and Assessment as tools for Moroccan Higher Education
Governance and Management in partnership with the European Union.
Noreddine ABGHOUR is currently an Assistant Professor at the Faculty of
Sciences Ain Chok, Casablanca, Department of Mathematics and
Computer Science. He is also a PhD in Computer Science, area of
distributed applications, from The Polytechnic Institute of Toulouse. In
addition to teaching, his research interests are in the field of security and
collaborative systems.
