A Novel Tool to Predict the Impact of Adopting a Serious Game on a

Learning Process

Ibtissem Daoudi, Raoudha Chebil and Wided Lejouad Chaari

National School of Computer Sciences (ENSI), Manouba University (UMA), Manouba, Tunisia

Keywords:

E-learning Process, Emotional State, Multi-agent-based Simulator, Serious Game.

Abstract:

In recent years, the rapid development in information and communication technologies has provided the lear-

ning ﬁeld with a variety of new teaching methods to motivate students and improve their skills. Serious Game

(SG) is an example of these new forms of learning which constitutes an attractive way supposed to replace

the classical boring courses. However, the use of SGs in classroom teaching is still limited since the choice of

the adapted SG to a speciﬁc learning environment remains a challenging task that makes teachers unwilling to

adopt this concept.Face to this ﬁnding, our aim is to propose a multi-agent-based simulator to predict the ef-

fect of a SG adoption in a learning environment given several game and players characteristics. As results, the

simulator gives intensities of several emotional aspects characterizing learners reactions to the SG adoption.

Experimentation demonstrates that the results given by the proposed tool are close to real feedbacks. This

work is supposed to encourage the use of SGs by giving an expectation of its impact on e-learning processes.

1 INTRODUCTION AND

MOTIVATION

Nowadays, technological progress concerns practi-

cally all the domains. In learning ﬁeld, this pro-

gress has produced the concept of Serious Game (SG)

which offers more attractive and motivating learning

environments than classical teaching methods. SG

can be deﬁned as ”the game designed for a seri-

ous purpose other than pure entertainment” (Julian,

2007). Currently, a lot of SGs have been developed

and are available for putting into use. In the literature,

several works have been proposed to evaluate SGs in

real situations of learning using self-report question-

naires. They studied different evaluation criteria such

as motivation (Lotﬁ, 2013), Flow (Freitas et al., 2014)

and learners knowledge and/or competencies (Oul-

haci, 2014) (Daniel et al., 2015) (Thomas et al., 2012)

(Muratet et al., 2016).

Despite the multitude of evaluation works in SGs

and their interest, the exploitation of this concept is

still limited in learning processes (Xu and Frezza,

2011). In fact, many teachers are reluctant to take

these games into their courses because they don’t

know if the adoption will be a success or a failure,

and there is no evaluation results for them to refe-

rence. Moreover, the integration of a SG in a real lear-

ning process is a difﬁcult task (Martens and Mueller,

2016) and does not necessarily guarantee good outco-

mes. Indeed, when the chosen game is not adapted

to a speciﬁc learning environment, this can cause ne-

gative emotions among the learners such as boredom,

anxiety or abort. This fact will, in our opinion, af-

fect negatively the evolution of the learning process

as well as a huge waste in terms of time and money.

Hence, the choice of a SG must be carefully studied

before integrating it in a course by considering the

most important features of the learning environment

including the learner-player and the SG. As a solution

to this problem we propose, in this paper, a novel tool

for simulating the integration of a serious game in an

e-learning process. This tool allows teachers to pre-

dict the impact of a SG on their learning environments

before deciding to use it. The simulator receives as

input several characteristics of the game and the play-

ers, and thanks to speciﬁc functions, generates a re-

port to predict the results of the learning environment

and emotional states of learners.

The rest of this paper is structured as follows.

Section 2 overviews some background information on

SG. Section 3 positions the reader in the context by

summarizing most of the existing work on SGs evalu-

ation and simulation. Section 4 details the proposed

model of a serious game-based environment. Accor-

ding to this, section 5 describes the implementation

and results of our simulator. To verify the reliability

Daoudi, I., Chebil, R. and Lejouad Chaari, W.

A Novel Tool to Predict the Impact of Adopting a Serious Game on a Learning Process.

DOI: 10.5220/0006685905850592

In Proceedings of the 20th International Conference on Enterprise Information Systems (ICEIS 2018), pages 585-592

ISBN: 978-989-758-298-1

585

of the proposed tool, section 6 compares the results

given by the simulator to real feedbacks obtained after

adopting the SG in a real learning process and gives a

discussion of the case study. Finally, section 7 sums

up the conclusion and outlines future works.

2 BACKGROUND ON SERIOUS

GAME

A survey on the term ”serious game” showed that

there is no current singleton deﬁnition of the concept

and provided a wide range of deﬁnitions. One of the

most recent deﬁnitions of this term can be found in

the PhD thesis by (Julian, 2007). He deﬁned a seri-

ous game as: ”a computer application, which aims to

combine aspects of both serious as, but not limited to,

teaching, learning, communication, or further infor-

mation with entertainment from the spring game”.

Actually, SGs provide a more powerful means of kno-

wledge transfer which can be used in several domains

for different purposes such as education, ecology, mi-

litary, and health-care. In fact, there is two play modes

in SGs: multi-players and single-player. The multi-

players mode is a mode of play involving more than

one person at the same time in a shared game environ-

ment, whereas a single-player mode is a play mode

designed for involving only one person. Some of SGs

offer the two play modes and others are limited to

only one play mode. To be more concrete, Table 1

presents some examples of existing SGs.

Table 1: Examples of serious games.

Serious

Game

Domain Game Purpose Multi-

Players

Navadra Education Educative

message broad-

casting

Yes

804 Military Marketing

message broad-

casting

Rescue-

Sim

Crisis

Manage-

ment

Training Yes

Eye sur-

gery

Health-

care

Training No

Aquacity

Game

Ecology Informative

message broad-

casting

Flee the

skip

Corporate Communication

message broad-

casting

Yes

3 RELATED WORK

The state of the art of SGs is quite rich and the exis-

ting works can be classiﬁed into several categories ac-

cording to the problem type. Since we are interested

in simulating and evaluating serious game-based en-

vironments; we targeted works related to these topics.

Consequently, the considered works may be divided

into two main groups.

The ﬁrst group consists of the Non-Player Cha-

racter (NPC) simulation in SGs. For example, (Ochs

et al., 2009) as well as (Karim, 2014) have developed

tools that model and simulate the non-player behavior

in SGs to improve the NPC credibility and to increase

consequently the feeling of immersion and pleasure

among players.

The second group focuses on learners’ assessment

and evaluation during a game session. To attain this

goal, many works proposed different criteria such as

motivation, Flow and learners knowledge and/or com-

petencies. For instance, (Lotﬁ, 2013) focused on as-

sessing motivation basing on the motivational model

ARCS (Attention, Relevance, Conﬁdence and Satis-

faction) combined with electro-physiological recor-

dings since it represents a key factor of an efﬁcient

learning. In addition, (Freitas et al., 2014) conﬁrmed

the relevance of the use of Flow criterion to evaluate

the optimal learning experience with a SG. They con-

ﬁrmed also the effectiveness of using self-report que-

stionnaire as an evaluation method since it provided

results conform to real feedbacks. Furthermore, the

research works referenced by (Oulhaci, 2014) (Daniel

et al., 2015) (Thomas et al., 2012) and (Muratet et al.,

2016) focused on the learning aspect to evaluate the

knowledge and/or competencies acquisition during a

game session.

Let us remember that our aim is to simulate a

learning environment based on a SG by taking into

account the learner emotional state as well as diffe-

rent features representing the game. In the literature,

there is no works having exactly the same purpose.

On the one hand, (Ochs et al., 2009) as well as (Ka-

rim, 2014) allow us to learn about emotions which

constitute a crucial factor impacting the SG progress.

However, their objectives are different from ours. In

fact, we propose to develop a SG simulator allowing

teachers to predict the impact of adopting a SG on

a particular learning process by taking into account

the learner emotional state. On the other hand, the

evaluation works referenced by (Lotﬁ, 2013) (Frei-

tas et al., 2014) (Oulhaci, 2014) (Daniel et al., 2015)

(Thomas et al., 2012) and (Muratet et al., 2016) are

so important because they allow us to identify several

success criteria of a serious game-based learning ses-

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

586

sion. Whereas, they are all applicable after the SG use

and there is no work that allows to give a prediction

on learning effects using SGs.

This lack has motivated the current work aiming

to propose a tool for simulating the integration of a

serious game in an e-learning process. This tool al-

lows teachers to study the impact of a SG (in terms

of success degree) on their learning environments be-

fore deciding to use it. To be able to implement the

simulator, we started by proposing a model of SG en-

vironments which will be described in the following

section.

4 SERIOUS GAME-BASED

ENVIRONMENT MODELING

Since our aim is to predict the success (or failure) of

using a serious game in a learning process, the propo-

sed model is based on meaningful success features of

the considered environment. In what follows, we start

by describing the selected features related to learners

as well as the serious game. After that, we present

the functions linking these features to each other, and

ﬁnally we detail the proposed model.

4.1 Success Factors and Success

Indicators

Our contribution is inspired by studies focused on

criteria used to evaluate serious games (Lotﬁ, 2013)

(Freitas et al., 2014) (Daniel et al., 2015) (Calderon

and Ruiz, 2015). Thanks to the considered works, we

were able to extract two feature types consisting of:

”success indicators” and ”success factors”.

The considered success indicators are: interest de-

gree, immersion degree, motivation degree and Flow

degree. These indicators are supposed to indicate the

success degree of the game-based learning environ-

ment and are impacted by success factors. The con-

sidered success factors are: the game context realism,

the gameplay, the game relevance and conﬁdence, the

game attention, the game challenge and the player

skills. In the proposed simulator, the success factors

will represent the input and the success indicators will

represent the output.

For more clarity, the deﬁnition of these criteria is

given in Table 2.

In the following part, we describe the different

functions linking success factors with success indica-

tors.

Table 2: Modeling Features Deﬁnition.

Feature name Deﬁnition

Success factors:

Game challenge The game difﬁculties and

problems.

Game context rea-

lism

The usefulness of game con-

tent in real life.

Gameplay The ability of the game to be

played.

Game attention The presentation style of

game content.

Game relevance

and conﬁdence

The importance and the ease

of learning.

Player skills The player aptitudes, capa-

cities and abilities.

Success indicators:

Flow degree The feeling of being pleased

to play game.

Immersion degree The feeling of being invol-

ved in game.

Motivation degree The feeling of being motiva-

ted to play game.

Interest degree The feeling of giving con-

cern to game content.

4.2 Links between Success Factors and

Success Indicators

4.2.1 The Interest Degree

The game context realism expresses the ability of the

serious game to describe real situations and concrete

scenarios that can be applicable in real life. The fee-

ling of interest is strongly linked to the context rea-

lism. In fact, when the context realism level increases,

the interest becomes more and more intense. (Hidi

and Renninger, 2006) distinguishes three interest de-

grees: the triggered situational interest referring to

a psychological state resulting from an interest to a

game-based learning process; the maintained situa-

tional interest referring to a psychological state that

contributes to the maintenance of situational interest;

and the individual interest characterized by more

stored knowledge and more stored value for particular

content than for other activity.

The proposed function, as shown in Figure 1, is

inspired by ﬁve-point Likert scale which is one of the

most popular psychometric scale used to measure so-

meone’s attitudes or behaviors. The passage from one

state to another is conditioned by the realism of con-

text level value. For example, the passage from the

triggered situational interest to the maintained situa-

tional interest is assured if the assigned value to the

context realism is equal to 3. Similarly, we attain the

A Novel Tool to Predict the Impact of Adopting a Serious Game on a Learning Process

587

individual interest when the provided value is equal to

the maximum value 5.

Figure 1: Evolution of Interest Degree Depending on the

Context Realism Level.

4.2.2 The Immersion Degree

The immersion degree of the learner in the virtual

word of the game is strongly impacted by the ga-

meplay aspect. The gameplay aspect expresses the

ability of the game to be played and includes several

factors like sound elements, animations and graphi-

cal quality (St-Pierre, 2010). We propose three diffe-

rent immersion degrees: the interaction taking place

while reading the rules at the beginning of the game

session; the engagement happening when the player

is actively involved in the resolution of a particu-

lar problem; and the immersion detected when the

player is so involved in the game world that he beco-

mes unconscious of the time and himself.

As shown in Figure 2, we propose in the same way

a function containing a scale of ﬁve points to represent

the evolution of the immersion degree according to

the gameplay level.

Figure 2: Evolution of Immersion Degree Depending on the

Gameplay Level.

4.2.3 The Flow Degree

The most important concept used to explain sub-

jective experience while playing games is Flow the-

ory. (Csikszentmihalyi, 1990) deﬁned the Flow as:

”rewarding, subjective, emotional state of optimal

pleasure that arises when an individual is absorbed

in either work or leisure activities that are perceived

as valuable” . This state depends on the actor skills

and the activity challenge. In fact, the Flow state

occurs when there is a perception of a balance bet-

ween the skill level and the challenge level as shown

in Figure 3. Boredom and anxiety are negative ex-

periences that demotivate the player: if the player is

bored, he has to increase the challenge he is facing. In

contrast, if the player feels anxiety, he must increase

his skills. Apathy is an emotion that occurs when the

values of skill level and challenge level are equal but

they are not maximal.

Figure 3: Relation between the Skill Level and the Chal-

lenge Level (Csikszentmihalyi, 1990).

4.2.4 The Motivation Degree

It is commonly accepted that ”motivation” plays a

huge role in learning, further suggesting that hu-

man beings learn best when having fun (Martens and

Mueller, 2016). Basing on the motivational model

ARCS (Keller, 2010), we note that the motivation le-

vel is strongly impacted by two game characteristics

including the attention level as well as the relevance

and conﬁdence level. In fact, the attention level ex-

presses the fact to attract the player attention on the

presentation style of a game content. The relevance

and conﬁdence level expresses the perception of the

importance and the ease of learning. Indeed, when the

player believes that the level of relevance and conﬁ-

dence is high and that the attention level is low, a dis-

couragement feeling will be occurred. Contrarily, a

relevance and conﬁdence perceived as lower than the

attention level will be a source of indifference as il-

lustrated in Figure 4. The feeling of motivation is

occurred when there is a perception of a balance bet-

ween the attention level and the relevance and conﬁ-

dence level.

The previously discussed aspects are summarized

in Table 3: for each success indicator, the correspon-

ding degrees as well as the success factors impacting

it are shown.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

588

Figure 4: Relation between the Attention Level and the Re-

levance and Conﬁdence Level.

Table 3: Modeling features in serious games.

Success factors Success

indicators

Corresponding

degrees

Game context re-

alism

Interest

degree

(Hidi and

Renninger,

2006)

Triggered situ-

ational interest,

maintained situ-

ational interest,

individual inte-

rest

Gameplay Immersion

degree

(St-Pierre,

2010)

Interaction,

engagement,

immersion

Game attention Motivation

degree

(Keller,

2010)

Indifference,

discourage-

ment, motiva-

tion

Game relevance

and conﬁdence

Game challenge Flow

degree

(Csik-

szent-

mihalyi,

1990)

Boredom, anx-

iety, apathy,

Flow

Player skills

4.3 The Proposed Success-Oriented

Model

In a multi-players SG environment, players control,

via their keyboard and their mouse, their Player Cha-

racter (PC) within the game graphical interface. The

PC evolves in the game environment including enti-

ties and objects populating it and interacts with ot-

her PC and Non-Player Character (NPC is a charac-

ter controlled by the game artiﬁcial intelligence). The

player reactions to the changes produced in a game

environment depend on his skills level, skills of ot-

her players as well as the characteristics of the game

consisting of: the challenge level, the gameplay le-

vel, the relevance and conﬁdence level, the attention

level and the context realism level. The PC or the

NPC can be represented by a software agent who is

able to perceive other players characteristics and to

make decisions basing on his information in order to

adapt his characteristics to the considered game envi-

ronment (Daoudi et al., 2017).

Basing on this brief description, we propose the

success-oriented model of serious game-based envi-

ronments shown in Figure 5.

Figure 5: The Proposed Success-Oriented Model.

5 SIMULATION DETAILS AND

RESULTS

In this section, we focus on the simulation work. For

this aim, we start by describing the simulator imple-

mentation, after that we present the simulation pro-

cess, and ﬁnally we speak about simulation results.

5.1 Simulator Implementation

A learning environment based on a SG is constitu-

ted by human and non-human players who have the

capacities of perception and action. For this reason,

our simulator is represented by a multi-agent system

which is a set of agents interacting with each other,

situated in a common environment and able to control

their own behavior according to their own goals. In

this work, our simulator is limited to represent a SG

environment consisting in different reactive agents in-

teracting with a game since they just have reﬂexes

A Novel Tool to Predict the Impact of Adopting a Serious Game on a Learning Process

589

without maintaining any internal state. Since the pro-

posed model is based on the representation of human

and non-human players, the resulting simulator can be

used to simulate both multi-players and single-player

SGs. In fact, our system implemented, thanks to JAVA

Agent DEvelopment Framework, four types of agents

with different roles. In the following points, we des-

cribe the role of each agent.

• Player Agent: this agent represents the learner-

player who is characterized by a skill level. His

role consists in perceiving the others agents cha-

racteristics to update his skill level and act on the

game environment.

• PC Observer Agent: this agent calculates the

average skills of all player agents. He informs

them periodically about this value in order to up-

date the challenge value perceived by each agent

player.

• NPC Observer Agent: the role of this agent is to

inform all the player agents about the non-player

agents skill levels.

• Non-player Agent: this agent represents the NPC

of the game who is characterized by a ranking.

This ranking is identiﬁed according to the difﬁ-

culty degree of the game selected by the teacher

using the simulator interface. This ranking is sent

to the NPC observer agent to inform it about the

skill level of the considered NPC.

5.2 Simulation Process

In order to clarify the followed approach, we describe,

in this part, the simulation process. As illustrated in

Figure 7, the main steps of the simulation process are:

the data collection, the simulation execution and the

results display. In the following, we detail each step

of this process:

1. Data Collection: this ﬁrst step consists in col-

lecting data about the game environment through

a questionnaire. This questionnaire is intended to

teachers who wish to simulate a serious game be-

fore integrating it in a particular learning process.

The teacher must quantify the game characteris-

tics consisting of: the challenge level, the rele-

vance and conﬁdence level, the gameplay level,

the context realism level and the attention level.

2. Simulation Execution: Statistics on the questi-

onnaire responses provide numerical values repre-

senting the intensities of the previously cited fea-

tures. These values are considered as the simu-

lator inputs; their availability allows to start the

simulation process. There are also other features

which are collected through a graphical interface

like the percentages of players having a skill level

x (x between 1 and 5) and the game difﬁculty to

determine the NPC skill level as shown in Figure

6. From the input values and according to the pro-

posed model, PC and NPC are created to simulate

the game. Thanks to speciﬁc functions connecting

several game environment features, the values of

these features are periodically updated.

Figure 6: The Simulator Interface.

3. Results Display: this step consists in displaying

the generated results in a graphical form.

Figure 7: The Simulation Process.

5.3 Simulation Results

In order to present and to explain the form of the si-

mulator results, we rely on the graphical interfaces

obtained after the simulation of the SG ”CodeCom-

bat” which will be described in the following section.

The simulator gives two forms of results: the ﬁrst dis-

plays the individual emotions felt by each player as

shown in Figure 8 and the second shows a global view

of the felt emotions as illustrated in Figure 9.

Figure 8 depicts the emotion intensities as percen-

tages: it shows that the player identiﬁed by ”PC2”

feels 25% of boredom, 100% of engagement, 25% of

indifference and 100% of triggered situational inte-

rest. These emotions represent respectively the Flow

degree, the immersion degree, the motivation degree

and the interest degree.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

590

Figure 8: Emotions Felt by the Player ”PC2”.

Figure 9: Percentages of Global Emotions.

Figure 9 summarizes the global emotional ra-

tes. In fact, the simulation of the SG ”CodeCombat”

shows that it is foreseeable to have: 60% of play-

ers feeling boredom, 30% expressing apathy and 10%

feeling anxiety. About the interest degree, it is en-

visaged that 50% of learners feel triggered situational

interest, 40% feel maintained situational interest and

10% express individual interest. Besides, the report

indicates that it is expected to have: 60% of players

feeling motivation, 30% expressing indifference and

10% feeling discouragement. Moreover, the study

shows, as a predictable result, that 10% of learners

feel interaction, 80% feel engagement and 10% ex-

press immersion.

The generated results aim to help teachers to pre-

dict the impact of adopting the game ”CodeCombat”

in a particular learning process by analyzing the re-

sulting learners emotions. These values are interes-

ting to the extent that they are consistent with the re-

ality. In the next section, we discuss our ﬁndings by

comparing the simulator results to real results.

6 DISCUSSION

In order to verify the reliability of the proposed si-

mulator, this section aims to compare the emotional

predictions to real feedbacks. The experimentation is

carried out in the context of a programming course

in an engineering school. It is based on the SG ”Co-

deCombat” which is a multi-players game designed

for learning programming languages. Given the pur-

pose of this experimentation work, it is composed by

two parallel steps: a simulation of the considered le-

arning process based on the proposed tool and a real

performing of a programming learning session based

on ”CodeCombat”. Once the two steps are achieved,

their results are compared. After the learning session,

we proposed a specially designed questionnaire to all

the participants in order to report the emotions felt by

each player and to verify their conformity to the si-

mulator results. To perform the simulation step, the

responsible teacher of the previously described lear-

ning session replied to the questionnaire described in

Section 5. He also gave the game features and laun-

ched the simulation using the simulator interface. Fi-

gure 10 summarizes the results obtained from the ex-

perimentation and the simulation process.

Figure 10: Comparative Chart of Simulator Results and

Real Results.

As shown in Figure 10, the simulator gave results

close to the reality, which is considered as a positive

outcome. After the learning session, we noted that

the use of a SG to learn how to code was, for the ma-

jority of learners, a good experience (57.2% felt mo-

tivation): the student discovered a modern learning

tool more motivating than the classical boring cour-

ses. However, the players feedbacks showed that the

game created boredom and there is no student who at-

tained the ﬂow state (0%). This ﬁnding is explained

by the fact that the game challenge level is not appro-

priate to the skill level of learners. We noted also that

the rate of individual interest is low (7.1%) which me-

ans that the game content does not attract players cu-

riosity to learn new coding techniques. So, the game

A Novel Tool to Predict the Impact of Adopting a Serious Game on a Learning Process

591

must propose more concrete scenarios. Concerning

the immersion degree, we found that 7.1% of learners

felt immersed which proves that the gameplay aspect

is not well designed and should be more attractive in

terms of graphical interfaces.

7 CONCLUSION AND FUTURE

WORK

In this paper, we presented the results of investiga-

tions on an important challenge consisting in serious

games adoption in learning processes. The paper aims

to develop a tool to simulate the use of a SG before

integrating it in a particular learning process. For this

purpose, we proposed a success-oriented model ba-

sed on emotional states of learners and different fe-

atures of the SG. Then, we developed a multi-agent-

based simulator which would be able to predict the

impact of operating a SG on classroom teaching. On

the other hand, the research got the result that came

from the real operation by studying the playing ex-

periences of the SG ”CodeCombat”. After that, we

conducted a comparison of the simulator results with

experimental results in terms of global emotional ra-

tes. Our ﬁndings show that the simulator gives results

close to real feedbacks. The proposed tool is intended

to teachers wishing to integrate a SG in their classical

courses. It allows them to study the adequacy of the

SG to the skill level of their students. So, this novel

tool is able to encourage the passage from traditional

to modern learning methods by giving an expectation

of the effect of using SGs in a learning process.

As an immediate future work, we aim to extend

the proposed simulator by considering other success

indicators and factors. Furthermore, one of our future

aims is to validate the simulator in different game con-

texts with other student populations.

REFERENCES

Calderon, A. and Ruiz, M. (2015). A systematic literature

review on serious games evaluation: An application to

software project management. Computers & Educa-

tion, 87:396–422.

Csikszentmihalyi, M. (1990). Flow: The psychology of op-

timal experience. Harper & Row.

Daniel, L., Olfa, C., and Imed, B. (2015). Retour

d’experience sur l’insertion d’un serious game dans

l’apprentissage des systemes d’information. Ingenie-

rie des Systemes d’Information, 20(1):11–36.

Daoudi, I., Chebil, R., and Chaari, W. L. (2017). A multi-

agent simulation of serious games to predict their im-

pact on e-learning processes. International Journal of

Social, Behavioral, Educational, Economic, Business

and Industrial Engineering, 11:405–413.

Freitas, S. D., Arnab, S., kiili, K., and Lainema, T. (2014).

Flow framework for analyzing the quality of educatio-

nal games. Entertainment Computing, 5(4):367–377.

Hidi, S. and Renninger, A. (2006). The four-phase model

of interest development. Educational Psychologist, 41

(2):111–127.

Julian, A. (2007). Du Jeu video au Serious Game: appro-

ches culturelle, pragmatique et formelle. PhD thesis,

Toulouse University II, Toulouse University III.

Karim, S. (2014). Adaptation dynamique des Environne-

ments Informatiques pour l’Apprentissage Humain.

PhD thesis, Laboratoire d’Informatique en Images et

Systemes d’information.

Keller, J. (2010). Motivational Design for Learning and

Performance: The ARCS Model Approach. Springer.

Lotﬁ, D. (2013). Contribution de la motivation dans les

jeux serieux. PhD thesis, Montreal University.

Martens, A. and Mueller, W. (2016). Gamiﬁcation - a struc-

tured analysis. IEEE 16th International Conference

on Advanced Learning Technologies, pages 138–142.

Muratet, M., Yessad, A., and Carron, T. (2016). Frame-

work for learner assessment in learning games. In

11th European Conference on Technology Enhanced

Learning.

Ochs, M., Sabouret, N., and Corruble, V. (2009). Simu-

lation de la dynamique des emotions et des relations

sociales de personnages virtuels. Revue des Sciences

et Technologies de l’Information, 23:327–357.

Oulhaci, M. A. (2014). Evaluation individuelle et collective

dans les jeux serieux collaboratifs: Application a la

gestion de crise. PhD thesis, Aix-Marseille Univer-

site.

St-Pierre, R. (2010). Des jeux video pour l’apprentissage?

facteurs de motivation et de jouabilite issus du game

design. DistanceS, 12(1):4–26.

Thomas, P., Labat, J.-M., Muratet, M., and Yessad, A.

(2012). How to evaluate competencies in game-based

learning systems automatically? In Proceedings of the

11th International Conference on Intelligent Tutoring

Systems, pages 168–173.

Xu, W. and Frezza, S. (2011). A case study: Integrating a

game application-driven approach and social collabo-

rations into software engineering education. In Pro-

ceedings of the 13th International Conference on En-

terprise Information Systems.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

592