FROM AN E-LEARNING TO AN U-LEARNING ENVIRONMENT

Ana Marilza Pernas

1,2,3

, Isabela Gasparini

1,4

, Amel Bouzeghoub

, Marcelo Pimenta

Leandro Krug Wives

and José Palazzo Moreira de Oliveira

Instituto de Informática, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil

TELECOM & Management SudParis, France

Universidade Federal de Pelotas, Brazil

Universidade do Estado de Santa Catarina, Brazil

Keywords: AdaptWeb, u-Learning, Situation-awareness, Context-awareness, User model.

Abstract: AdaptWeb

is an e-learning environment designed to offer personalized content to different groups of stu-

dents according to student’s models that are mainly based on personal information. This paper shows how

the platform is modified to support ubiquitous learning, managing mobility, social interaction, device inde-

pendence, and context awareness in a broader and richer notion of the student’s context. The paper presents

the architecture and its implementation by three different kinds of data servers, working together for model-

ing the situations experienced by the students, storing their profiles and contexts, and maintaining corres-

ponding learning objects.

1 INTRODUCTION

Learning in the traditional class model has been

viewed as “one size fits all” approach; all students

receive the same content without regard with the

student’s different background, preferences, skills

and learning styles. Such diversity has brought new

opportunities and challenges; one of the most de-

sired challenges is to present an adaptive and perso-

nalized behavior (Brusilovsky and Peylo, 2003).

Personalization may be characterized as the process

of adapting a computer application to the needs of

specific users, taking advantage of users’ behavior

analysis. An ELE, however, may be dynamically

adjusted not only according to the student’s model

but also depending on a richer notion of context

(Eyharabide et al., 2009). In our approach, ‘context’

is related to all resources that influence and are in-

fluenced by the student during a particular situation.

That situation may be any learning activity like

solving a list of exercises, answering an evaluation

or attending to a class. Context is related to the dif-

ferent resources that surround the student perform-

ing a task, e.g., location (rooms, buildings, and labo-

ratories), available devices, people, and available

learning objects. A contextualized ELE provides the

student with exactly the needed material, appropriate

to his/her knowledge level and suited for a special

learning situation called scenario (Eyharabide et al.,

2009). We assume that the inclusion of the situation

experienced by each student can improve the learn-

ing process (Bouzeghoub et al. (2007) and Ogata

and Yano (2004)).

This paper presents an extension to an e-learning

environment – AdaptWeb

(Adaptive Web-based

Learning Environment), focusing on how its struc-

ture and contents could be modified to support ubiq-

uitous learning (u-learning). In this new approach,

three servers are responsible for modelling the situa-

tions experienced by the students, storing their pro-

files and contexts, and maintaining corresponding

learning objects. To support this approach, a con-

text-aware architecture was defined, in which ontol-

ogy-based and scenario-oriented aspects are com-

bined to explicitly represent rich context, supporting

the context and scenario concepts as an extension to

traditional user-student modelling. This paper is

structured as follows: next section gives an overview

of the main concepts. Section 3 summarizes relevant

related works. Section 4 briefly describes the

AdaptWeb

environment. Section 5 explains our

extension, illustrating its behaviour by the means of

some context-aware scenarios. Finally, Section 6

presents concluding remarks.

180

Pernas A., Gasparini I., Bouzeghoub A., Pimenta M., Krug Wives L. and Palazzo Moreira de Oliveira J. (2010).

FROM AN E-LEARNING TO AN U-LEARNING ENVIRONMENT.

In Proceedings of the 2nd International Conference on Computer Supported Education, pages 180-185

DOI: 10.5220/0002774501800185

 SciTePress

2 FUNDAMENTALS

The research and practice in Computer Science and

Education has evolved with the introduction of the

Internet and Web-based courses. However, some

educational applications are usually developed with-

out taking into account the dynamic capabilities and

personalization that the Web environment can pro-

vide. This limitation raises serious usability issues

(Kalbach, 2007) such as (i) guidance problems, i.e.,

pages always presenting the same fixed content, and

(ii) navigation problems, with are the consequence

of a totally opened linking schema. A ubiquitous

learning environment (ULE) combines the advan-

tages of adaptive learning, the benefits of ubiquitous

computing and the flexibility of mobile computing

(Jones and Jo, 2004). U-learning enables the en-

hanced possibilities of accessing learning content

(typically learning objects) and computer-supported

collaborative learning environments at the right

time, at the right place, and in the right form (Boms-

dorf, 2005).

Like most of ELEs, AdaptWeb

employs student

modelling as a basis for the personalization process,

and the most common student’s models for e-

learning include knowledge about the student,

his/her background, learning styles and so on. In

order to provide such adaptation an ULE may be

dynamically adjusted not only according to the stu-

dent’s model but also depending on a richer notion

of context. In the present approach, we consider as

contextual information the student location, connec-

tivity, device, time, schedule, objectives, profile (e.g.

learning styles, interaction preferences), and learning

objects from the application domain, thus character-

izing a situation-aware environment.

In a situation-aware environment, a situation is

defined as the set of contextual characteristics that

are invariable in a defined time interval (Weißenberg

et al., 2006). Like Bouzeghoub et al. (2007) and

Yang et al. (2006), we also consider a situation as

consisting of a set of contexts over a period of time

that affects the future system behaviour. Different

types of events (Bouzeghoub et al., 2007) can be

defined in a situation-aware environment, and de-

pending on them a new situation may exist.

3 RELATED WORK

Although u-learning is a recent area, there are yet

many efforts focused on the development of ULEs.

An example is the UBI-Learn project (Laroussi and

Derycke, 2004) aiming to design a complex learning

dispositive, taking into account ubiquity and mobil-

ity. That project is mainly focused in the student

mobility, allowing them to interact in an augmented

class. Another ULE-related work is the CLUE pro-

ject (Ogata and Yano, 2004), which is a computer-

supported collaborative learning (CSCL) in a ubiqui-

tous environment. Their focus is a knowledge

awareness map that retrieves past interactions and

experiences based on the current learning context of

the student, providing the right information at the

right place.

In Sieg et al. (2007) is presented a framework to

integrate critical elements that make up the user

context, namely the user short-term behaviour, se-

mantic knowledge from ontologies that provide

explicit representations of the domain of interest,

and long-term user profiles revealing interests and

trends. They present an interesting approach for

building ontological user profiles by assigning inter-

est scores to existing concepts in the domain ontol-

ogy. They also present a framework for contextual

information access using ontologies and demonstrate

that this knowledge (combined with long-term user

profiles) can be used to tailor search results based on

users’ interests and preferences. The AdaptWeb

also aims to provide adaptation related to the user

location (Laroussi and Derycke, 2004), necessities

(Ogata and Yano, 2004) or behaviour (Sieg et al.,

2007), but combining the three forms of adaptation

in an integrated way.

4 ADAPTWEB

ENVIRONMENT

AdaptWeb

is an adaptive learning system whose

purpose was to adapt content, presentation and navi-

gation in an educational web course. Currently, it is

an open source environment fully operational and

actually being used on different universities. It is

available at SourceForge.

AdaptWeb

is composed by an authoring envi-

ronment where the teacher (author) organizes and

creates the content structure of their disciplines, and

by a students’ environment that shows personalized

views of the content, interface and navigation to the

student. The educational content of AdaptWeb

modelled through a hierarchical structure of con-

cepts where the criteria of prerequisites are estab-

lished. This structure is defined during the author-

ship’s phase and then stored in XML. The XML

documents go then through a filtering process,

which happens dynamically as the student interacts

with the environment.

FROM AN E-LEARNING TO AN U-LEARNING ENVIRONMENT

181

The student’s model describes users in terms of

theirs characteristics as knowledge, interactions

preferences, background, technological resource,

navigational history and cognitive learning styles.

The user’s background analyzes the student occupa-

tion. For instance, a Computer Science student and a

Mathematic student need distinct content of the logic

discipline. The teacher, through his experience,

determines which content will be provided for each

group of students (knowledge depth), thus enabling

the content to be adapted for each student.

User’s interaction preference environment sets

two navigational conducts for a discipline: guided

tour or free mode. Guided tour takes into account the

prerequisites set by the author at the stage of author-

ship, and the student can only access a concept if its

prerequisites are already known. In a free mode, the

student can navigate throughout the contents of a

discipline without restrictions. The student’s knowl-

edge is supposed to improve during the navigation in

a course and then the environment monitors the

student’s navigation, making constant updates in the

user profile. Finally, we assume that the learning

process will be improved if the content is accessible

in several media formats, including video, adapted to

different formats and resolutions. While addressing

these variation possibilities, the system allows the

delivery of videos with variable resolution and qual-

ity, according to the network bandwidth available.

5 TOWARDS U-LEARNING

In this section, we present how the AdaptWeb

architecture was extended to support mobility, social

interaction, device independence and context aware-

ness, being adapted to specific scenarios. First, we

have restructured the existing AdaptWeb

architec-

ture to be Service-Oriented (SOA), having different

data and context modules that are individually

adapted and orchestrated by a set of high level ser-

vices, allowing our environment to communicate

with other ones in a flexible manner.

In this approach, the new architecture is based on

three servers that operate together to provide and

manage contextualized data according to the stu-

dent's scenarios. Each server manages specific data

related to the user context, being respectively re-

sponsible for the storage and adaptation of (i) infor-

mation about students (personal data, preferences,

objectives, knowledge background, behavior, learn-

ing styles, cultural context, etc.), (ii) environmental

context (information related to the user environment,

tasks, activities, time interval, devices, location), and

(iii) learning object’s information (documents pro-

vided by the educational environment to its users for

their learning).

However, since this information is managed and

stored separately and the context needed to the adap-

tation environment is diverse, it is orchestrated by an

internal component called Context Management

Service (Figure 1). This service is responsible for

analyzing the context managed by the servers, gene-

rating different scenarios that can be experienced by

the students in a specific period. These scenarios are

used to guide the adaptation (in the Adaptation En-

gine), and materialized in the interface rendered to

the user.

The main goals of the architecture are: (i) easily

reuse of educational resources, since they will be

adapted to the user scenario while the stored content

remains the same, (ii) integration into the existing

architecture, since the new architecture is supposed

to take advantage of the existing functionalities and

(iii) extensibility to other educational systems, using

standard technologies. The personalization is possi-

ble with the combination of contextual data related

to whom and where the user is, what he/she is doing

and what does he/she needs to achieve his/her edu-

cational targets.

5.1 Extended Architecture

The AdaptWeb

architecture (Figure 1) works as

follows. The User Interface Component module

interacts directly with the user (represented by data

flows ‘a’ and ‘b’), identified after a login procedure.

User id data, in conjunction with the current user

location and time, is sent to the Context Detec-

tor/Collector (‘c’), who sends this new information

to the servers (‘d’) and informs the Context Man-

agement Service that new scenarios must be con-

structed (‘e’), since new data is available in the serv-

ers (‘e’). The communication between the Context

Management Service and the servers is performed

through Web Services that are responsible for trans-

porting XML’s files in/outside the environment (‘g’

and ‘f’).

After obtaining student’s data, each server inter-

nally updates their models, i.e., the situation model

server uses information about tasks, activities, time,

location, and device to (re)structure generic adapta-

tion situations in the environment; the user model

server manages the student profile (containing in-

formation like background, interaction preferences,

learning style, cultural context, and other aspects

related to the students learning); the learning ob-

jects' model server uses data related to user discip-

CSEDU 2010 - 2nd International Conference on Computer Supported Education

182

lines and didactic meetings to (re)organize the learn-

ing objects according to his/her needs, adapted to the

context.

Figure 1: ULE-oriented architecture.

Therefore, the Context Management Service

communicates with the servers (‘f’ and ‘g’) to dy-

namically build student adaptation scenarios, which

are described by the tuple {situation, user profile,

learning objects}. This module communicates with

the Context Collector/Detector to identify when new

important data is included in the servers and when a

new scenario must be constructed. The Adaptation

Engine is responsible for the acquisition of the di-

verse scenarios provided by the Context Manage-

ment Service (‘h’) and for the ordering of these

scenarios to be adapted to the student. In other

words, it has to organize the user’s scenarios, to give

priority to some actions and perform changes in the

User Interface Component, personalizing the naviga-

tion scheme and context view of the student (‘i’).

During the user interaction with the system, actions

and events are collected by the Context Collec-

tor/Detector, which passes these data back to the

servers (‘d’). The servers update their models based

on these events. Thus, the environment keeps track

of the student actions, while being adapted to the

student’s needs. In the next subsections, the internal

functioning of each server is described.

5.1.1 Situation Model

The goal of the situation model is to define generic

clauses of situations, which can be combined, using

specific operators, to derive new generic situations.

Generic situations are those that could be applied to

different users or group of users. Accordingly to

activities, devices, time, and other variables valid for

a student in a specific time interval, different clauses

may be chosen to represent his/her situation.

The Situation Model comprises four ontologies:

activity; device; time and location. The activity

ontology contains information about the user’s activ-

ities, including classes, meetings, and their respec-

tive time interval (day, month, year, beginning and

ending times). This information helps the system to

identify which user’s activities are available in a

given period of time. The device ontology contains

information (display resolution, network bandwidth,

etc.) about the current user device, in order to allow

filtering information according to its technological

constrains. The time ontology is an abstract ontology

of time (Bouzeghoub, et al., 2009), like OWL-Time

(W3C, 2003). Finally, the location ontology contains

information about locations and places, which allow

the system to filter information according to its dis-

tance or proximity to the user. Beyond these ontolo-

gies, the Situation Model has some simple defined

situations, which can be addressed to specific stu-

dents, helping the user detection that is performed by

the Context Management Service. When the Context

Collector/Detector receives basic information related

to the tasks and activities selected by the user during

his/her first access in AdaptWeb

, it can send this

data to the situation server, which can define simple

situations that identify that user in the Situation

Model. For instance, if we want to state that “A

student is using a mobile device”, “He is taking a

specific course” and “He is studying for three

hours”, we could use the following expression in-

volving the user (student), device, activity and time

ontologies (Bouzeghoub, et al., 2009).

LearningSituation = {

use(O

User

.Student,O

Device

.MobDevice),

do(O

User

.Student,O

Activity

.Std),

haveSubject(O

activity

.Std, O

domain

.@Domain)

setDuration(O

activity

.Std,O

time

.@3h)}.

5.1.2 Student Model

An example of Student model is shown in Figure 2.

Figure 2: Student Ontology (Musa et al., 2004).

FROM AN E-LEARNING TO AN U-LEARNING ENVIRONMENT

183

In Figure 2, the central class st:Student has prop-

erties that characterize the student profile. The prop-

erty st:has LearningStyle indicates the student’s

cognitive learning style (Souto et al., 2002). The

property st:hasLearningGoal refers to an element of

awo:Course, containing customized discipline con-

tent for goals of a specific group of students. The

Boolean property st:wantsTutorial indicates if the

student prefers to work in a tutored mode (guided

tour) or not. Using an overlay model, the student’s

knowledge on each topic of the Knowledge Model is

indicated by the relation st:hasKnowledgeOn, i.e., if

an instance of the relation st:hasKnowledgeOn exists

relating the student to an instance of the class

awo:Topic, then the system believes the student has

knowledge about this topic. The property

awo:hasNetworkConnection indicates the kind of

Network Connection detected in the current session.

The property st:locationLearningTrajectoryWF

indicates the URL where the current learning trajec-

tory for the student is. The remaining classes in the

model represent the range of the properties defined

in the Knowledge Space Model (Musa et al., 2004).

We can also take into account the cultural con-

text, represented in practice by features that distin-

guish between the preferences of students from dif-

ferent regions, like languages, skills, values, etc.

(more details see Eyharabide et al. (2009)).

5.1.3 Learning Objects Model

The Domain Knowledge Ontology, located in the

learning object’s server, contains the knowledge

describing the pieces of content in the hyperspace. It

also contains the rules to assemble them correctly in

order to compute complex learning objects adequate

to the profile of each student. More details can be

found in Muñoz and Palazzo (2004).

5.2 Examples of Context Adaptation

To illustrate our approach, here we describe a few

examples (scenarios) of the improvements in perso-

nalization’s capabilities of AdaptWeb

. The illu-

strated scenarios are composed by different context

types. For each scenario, we present how the envi-

ronment should be personalized to the student.

In the first scenario, Smith is connected to a ULE

through a desktop computer inside located in his

university’s research laboratory. He is a Computer

Science student, and now he is considering to solve

a list of exercises given by his Calculus teacher. He

is not going well in the subject and, consequently, is

not obtaining satisfactory results. The user model

analyzes his number of mistakes and identifies that

he needs help resolving the exercises. At the same

time, the situation model can detect via the teacher’s

agenda that a chat with the students was previously

scheduled by the teacher to happen in 10 minutes.

These events detected by the situation model and

user model will start a service of notification in the

Context Management Service, informing that a

change of the current scenarios that are related with

these events may change. After a new orchestration

by the Context Management Service, the User Inter-

face Component can send a message to the student,

notifying him of this possibility to solve his doubts.

In a second scenario, Maria, a Portuguese-

speaking PhD student of Chemistry, has very good

skills in three different foreign languages (English,

Spanish and French). She is going to France, to

study for a short period of time – to work in a differ-

ent research group. When she gets there, the system

detects her location change and adapts her learning

objects and interface for the new linguistic environ-

ment where she is now inserted, selecting their

French version.

Finally, in third scenario, Smith (the same stu-

dent from the first scenario), will follow an exam in

a few days. The system detects this activity in his

agenda and organizes the learning objects in a priori-

ty order, indicating the most important ones for the

moment. The system also presents some exercises to

the student and indicates some complementary ma-

terial to complement the acquired knowledge about

the exam content.

These three examples show how our approach

can personalize the system to different scenarios

(user’s context, situation and learning objects). A

richer adaptation for a specific student generates a

better learning process, since all features are perso-

nalized. We are developing experimental work not

only to get overall feedback (mainly subjective)

from users but to statistically validate the approach.

6 CONCLUSIONS

The use of e-learning brought new focuses and re-

sulted in fundamental changes in teaching and learn-

ing. The ubiquitous computing defines a new para-

digm in the computing field, where the computing

capability is provided to users anytime, anywhere.

This approach offers new possibilities and new chal-

lenges! It is necessary to deal with context aware-

ness and its relationship with users. Context-aware

modeling extends traditional user modeling tech-

niques, by explicitly dealing with aspects that have a

CSEDU 2010 - 2nd International Conference on Computer Supported Education

184

significant influence on the learning process. It is

orchestrated by an internal component called Con-

text Management Service responsible for analyzing

the context managed by the servers, generating dif-

ferent scenarios that can be experienced by the stu-

dents in a specific time interval. This new adaptation

model is a must in providing a real application of the

semantic Web as proposed by Berners-Lee et al.

(2001). This approach is a real challenge to all Web

systems' developers; eight years after the scenarios

proposed in the Berners-Lee paper very few real

customizing agents are in operation. We put this

technology working in a real-life e-learning system.

In this paper, we have presented an extended ap-

proach to provide a richer personalization in the

AdaptWeb

environment, describing an extension of

its architecture, which is based on Web services

technologies and context-awareness, allowing the

adaptation of course content to the user’s context

and situation. As the learning environment is opera-

tional in some Universities the next research stage is

centered in gathering navigational and pedagogical

data from real classes to fine-tune the context-

scenarios modeling concept.

ACKNOWLEDGEMENTS

This work has been partially supported by Conselho

Nacional de Desenvolvimento Científico e Tec-

nológico - CNPq, Brazil, and by the projects Adapt-

SUR 022/07 (CAPES, Brazil) - 042/07 (Secyt, Ar-

gentina), and AdContext 547-07 (CAPES-

COFECUB).

REFERENCES

Berners-Lee, T., Hendler, J., and Lassila, O., 2001. The

Semantic Web: A New Form of Web Content That Is

Meaningful to Computers Will Unleash a Revolution

of New Possibilities, Scientific American, 284 (5), 28-

37.

Bomsdorf, B., 2005. Adaptation of Learning Spaces:

Supporting Ubiquitous Learning in Higher Distance

Education. Mobile Computing and Ambient Intelli-

gence: The Challenge of Multimedia, Dagstuhl Semi-

nar Proceedings.

Bouzeghoub, A., Do, K. N. and Lecocq, C., 2007. A Situa-

tion-Based Delivery of Learning Resources in Perva-

sive Learning. LNCS, Springer-Verlag, Berlin, Heidel-

berg.

Bouzeghoub, A., Wives, L.K. and Do, K.N., 2009. Situa-

tion-Aware Adaptive Recommendation to Assist Mo-

bile Users in a Campus Environment. In IEEE 23

In-

ternational Conference on Advanced Information

Networking and Applications, AINA. Bradford, UK.

Brusilovsky, P. and Peylo, C., 2003. Adaptive and intelli-

gent Web-based educational systems. In P. Brusi-

lovsky and C. Peylo (eds.), International Journal of

Artificial Intelligence in Education 13 (2-4), 159-172.

Eyharabide, V., Gasparini, I., Schiaffino, S., Pimenta, M.

and Amandi, A., 2009. Personalized e-learning envi-

ronments: considering students’ contexts. IFIP World

Conference on Computers in Education, v. 302, pp 48-

57, Springer.

Jones, V. and Jo, J., 2004. Ubiquitous Learning Environ-

ment: An Adaptive Teaching System using Ubiquitous

Technology. In R. Atkinson, C. McBeath, D. Beyond

the Comfort Zone: proceedings of the 21

ASCILITE

Conference, Perth, Western Australia.

Kalbach, J., 2007. Designing Web Navigation: Optimizing

the User Experience, O´Reilly.

Laroussi, M. and Derycke, A., 2004. New E-Learning

Services Based on Mobile and Ubiquitous Computing:

UBI-LEARN Project. Proc. of International Confe-

rence on Computer Aided Learning in Engineering

Education – CALIE, Grenoble, France.

Muñoz, L.S. and Palazzo, J. M. de O., 2004. Adaptive

Web-Based Courseware Development using Metadata

Standards and Ontologies. In International Conference

on Advanced Information Systems Engineering -

CAiSE, Riga, Latvia, pp. 414–428.

Musa, D.L., Muñoz, L.S. and Palazzo J. M. de O., 2004.

Sharing Learner Profile through an Ontology and Web

Services. In International Workshop on Database and

Expert Systems Applications Zaragoza, Spain, pp. 415-

419.

Ogata, H. and Yano, Y., 2004. Knowledge Awareness

Map for Computer-Supported Ubiquitous Language-

Learning. IEEE WMTE, Taiwan.

Sieg, A., Mobasher, B. and Burke, R., 2007. Representing

Context in Web Search with Ontological User Pro-

files. LNCS, Modeling and Using Context, v. 4635, pp

439-452, Springer.

Souto, M.A., Verdin, R., Wainer, R., Madeira, M., Warpe-

chowsky, M., Beschoren, K., Zanella, R., Correa, J.S.,

Vicari, R.M. and Palazzo, J. M. de O., 2002. Towards

an Adaptive Web Training Environmental Based on

cognitive Style of Learning: an Empirical Approach.

In International Conference on Adaptive Hypermedia

and Adaptive Web Based Systems, AH 2002. LNCS,

vol. 2347, Springer-Verlag, pp. 338-347.

Weißenberg, N., Gartmann, R. and Voisard, A., 2006. An

Ontology-Based Approach to Personalized Situation-

Aware Mobile Service Supply. Journal of GeoInfor-

matica, Springer, Netherlands.

W3C., 2006. Time ontology in OWL. W3C Working Draft,

Sept. Available in: http://www.w3.org/TR/owl-time/.

Yang, S. J. H., Huang, A. F. M., Chen, R. Shian-Shyong,

T. and Yen-Shih, S., 2006. Context Model and Con-

text Acquisition for Ubiquitous Context Access in U-

Learning Environments. IEEE International Confe-

rence on Sensor Networks, Ubiquitous, and Trustwor-

thy Computing.

FROM AN E-LEARNING TO AN U-LEARNING ENVIRONMENT

185