A Systematic Mapping on Machine Learning Algorithms and

Gamification Applied to Education

Vinicius Schultz Garcia da Luz

, Ezequiel Gueiber

1,2 b

, Simone Nasser Matos

Helyane Bronoski Borges

, Guataçara dos Santos Júnior

and Rui Pedro Lopes

Computer Science Dept., Federal University of Technology, Paraná (UTFPR), Ponta Grossa, PR, Brazil

Computer Science Dept., State University of Ponta Grossa (UEPG), Ponta Grossa, PR, Brazil

Research Center in Digitalization and Industrial Robotics, Instituto Politécnico de Bragança (IPB), Bragança, Portugal

Keywords: Machine Learning, Education, Games, Gamification, Mapping Review.

Abstract: Machine learning algorithms and gamification applied in educational environments promote more accurate

information gathering as students interact with games. They allow you to evaluate and analyse data from how

to improve gamified tools to stimulate teaching, retain student attention and interest, and optimize learning.

This paper has performed a systematic mapping to identify how machine learning algorithms are applied at

each game level.

1 INTRODUCTION

The use of games in educational environments

improves the learning process, encouraging students

to improve their skills in various areas, in addition to

shaping behaviour using techniques and elements in

environments not related to games (Kapp, 2012;

Lopes et al., 2018; Mora et al., 2017) as is the case

gamification. Gamification uses the dynamics and

mechanics of games as a tool to motivate and increase

commitment in the educational field (Cordero-Brito

and Mena, 2018) and promotes motivation and

engagement (Lee and Hammer, 2011).

Machine learning (ML) algorithms when used in

education games provides an automation of tools and

facilitates data analysis, in which playful experiences

can be measured through multidisciplinary

interactions (Cowley et al., 2014).

This paper summarized, by means of a systematic

mapping based on the method of Kitchenham and

Charters (2007), the state of the art of how the

machine learning and gamification algorithms are

being used in educational games levels. The

https://orcid.org/0000-0002-9616-3503

https://orcid.org/0000-0002-2993-0928

https://orcid.org/0000-0002-5362-2343

https://orcid.org/0000-0002-9153-3819

https://orcid.org/0000-0002-6234-7961

https://orcid.org/0000-0002-9170-5078

systematic mapping was carried out from 2003 to

2020 and used 5 digital repositories.

2 RELATED WORK

Considering the existing studies in the literature on

games or gamification in education, a common

practice is to carry out literature reviews in a more

systematic way. This contributes to assess the area of

research in which it is desired to propose a solution

not yet measured. To understand the application of

gamification in education, we sought to identify in the

literature the mapping or systematic review related to

games in education, and to select them by the subject

of gamification applied to education using the

following search strings: “education” and

“gamification” and “systematic review” or

“systematic mapping”. The works were searched in

the Elsevier (Science Direct), ACM, IEEEXplore,

Scopus and Google Scholar repositories.

Garcia da Luz, V., Gueiber, E., Matos, S., Borges, H., Santos Júnior, G. and Lopes, R.

A Systematic Mapping on Machine Learning Algorithms and Gamiﬁcation Applied to Education.

DOI: 10.5220/0010498103530361

In Proceedings of the 13th International Conference on Computer Supported Education (CSEDU 2021) - Volume 2, pages 353-361

ISBN: 978-989-758-502-9

353

132 studies were found and 14 of these were

removed because they were duplicated, leaving a total

of 121. Of these, 28 studies were related to review or

mapping involving: Gamification and Education. The

subjects addressed through the mappings or

systematic reviews found were analyzed to determine

their objective.

It was found that some papers were focused on

identifying how gamification is applied and at what

levels of education, such as: Morelock (2013),

Dicheva et al. (2015), Alanne (2016), Bodnar et al.

(2016), De Avila dos Santos and Luis Castro de

Freitas (2017), Ortiz-Rojas et al. (2017), Subhash and

Cudney (2018), Alhammad and Moreno (2018),

Souza et al. (2018), Borges et al. (2018), Cordero-

Brito and Mena (2018) and Gentry et al. (2019).

There are other works focused on the use of tools,

learning theories or environments in which

gamification is applied: De Sousa Borges et al.

(2014), Truong (2016), Stevenson et al. (2017), Brito

and Madeira (2017), Tenório et al. (2018), Trinidad

et al. (2018) and Lara et al. (2019). Other papers focus

on the review of how the gamification project is

developed and reused as described in Peixoto and

Silva (2015), Seaborn and Fels (2015), Mora et al.

(2017), Calderón et al. (2018) and Kamunya et al.

(2019). In addition to the systematic reviews that

address application, use of tools, learning theories,

environments in which gamification is applied and

projects for the development of gamified software

that was previously presented; there are also studies

on gamification and multidisciplinar and new

guidelines for the use of gamification, such as:

Osatuyi et al. (2018), Bozkurt and Durak (2018),

Rodrigues et al. (2019) and Inocencio (2018),

respectively.

In the research it was noted that there was an

increase in studies related to systematic mapping that

describe gamification in education (in 2018 there

were 7 papers). Analyzing the number of experiences

of gamification applied in educational levels, higher

education has a greater frequency with 6 works,

followed by 5 works in various levels (basic

education, high school and higher education), 16

without information and 1 in high school, elementary

school and teaching for the intellectual disabled.

The analysis of related works showed that higher

education concentrates most applications of

gamification in education in all years, and this

educational level also falls into several levels (basic

education, high school and higher education). In 2015

two studies were retrieved, one in higher education

and the other which does not reveal information on

which educational level the gamification context was

applied to. In addition, only in the year 2017 there are

publications of gamification applied to high school

and teaching for the intellectual disabled. An

important factor is that the student's development can

be carried out in an individualized or customized way,

knowing their difficulties, skills and potential (Mora

et al. 2017). It was found that the works address many

aspects related to gamification and games applied in

education, but do not show how machine learning

algorithms are applied at educational game levels.

3 REVIEW METHODOLOGY

Based on the literature, studies involving games in

education were found with the application of machine

learning algorithms. The method used in this work to

carry out systematic mapping was that of Kitchenham

and Charters (2007). The authors Martins and

Gorschek (2016) define that their work is a reference

for those who wish to carry out systematic reviews

and mappings; and report that this approach covers

several steps to generate comprehensive research.

The systematic mapping was carried out by the

authors of this work: 1 in the area of software

engineering, 1 in the area of artificial intelligence, 1

in the area of education and 1 in the area of

technology and management, 1 doctoral student in

Science, Technology and Teaching and 1 master's

student in Computer Science. It includes studies from

2003 to 2020 on the subjects of games and machine

learning algorithms aimed at education. The proposed

work carried out the research contemplating games,

as there could be works containing gamification

within the process of this systematic mapping.

At first, as predicted by the systematic mapping

method, the protocol was defined by the authors in

order to bring to the researcher a quantity of

information about: authors of works, relevance of

each work and their real contribution to the

community and research problems. In this protocol,

research questions, search bases and their specific

forms of search, search strings, keywords and filters

were defined, presented in the next sections. Figure 1

shows how the process of systematic mapping was

conducted. In the planning process, the objective of

systematic mapping was identified to verify the

educational games that apply machine algorithms in

order to check how it can help for a sustainable

education. The definition of the protocol used an

electronic spreadsheet for filling in information on the

work to be analyzed. In the selection of studies, the

search bases, search strings were defined and

CSEDU 2021 - 13th International Conference on Computer Supported Education

354

questions were elaborated to be answered in the

systematic mapping.

Figure 1: Research Methodology Adopted.

Then, the inclusion and exclusion criteria were

defined to select the desired works and form the

protocol with information from the articles so that the

questions could be answered. With the protocol

formed it was possible to synthesize the data, present

the results and verify future trends on the subject.

3.1 Research Questions

The central objective of the research was to list the

main tools and methods used in the researched

subjects. The following questions were elaborated:

QR1) What are the games styles, the area and the

level of education (special or not)?

QR2) How the games architecture is built?

QR3) What kind of machine learning algorithms and

techniques are used in the games?

3.2 Inclusion and Exclusion Criteria

The researchers who performed the systematic

mapping were the evaluators of the collected primary

works. Initially, an inclusion criterion ware studies

that presented in their Title, Keywords or Abstract

information related to the research topic and that were

within the time period determined for systematic

mapping.

The application of the inclusion and exclusion

criteria followed the steps: i) Reading the titles,

keywords and abstract, excluding works that are not

related to the inclusion criteria; ii) Dynamic

(superficial) reading of the study considering the

inclusion and exclusion criteria and iii)

Documentation of each study in short paragraphs.

Five search databases were defined to perform the

researches: Google Scholar, Science Direct, ACM,

IEEEXplore and Scopus. The choice of the cited

bases was because they had a greater number of

publications related to this work's research.

3.3 Search Strings and Keywords

Several search terms had already been thought and

discussed among the researchers, which facilitated

the process of composing the keywords. After some

tests, expressions restricted to the theme were used,

such as "artificial intelligence algorithm" AND

"gaming" AND education", "machine learning" AND

"digital games", "childhood" AND "education" AND

"machine learning" AND "games", "Machine

learning algorithm" AND "education games" AND

"artificial intelligence", "machine learning algorithm"

AND "education" AND "digital games", "machine

learning algorithm" AND "education" AND "games".

Using the before mentioned strings the number of

results was low or in some cases no results were

returned. With that, a refinement of the search string

was carried out, and the combination that returned the

most results was "machine learning algorithm" AND

"education" AND "games" to obtain publications of

systematic mapping articles. A total of 382 articles

were found using a set of keywords defined for

searching works related to games in education

applying machine learning algorithms. Table 1 shows

the number of works obtained per repository.

However, 13 duplicate works were eliminated: 3 from

the Scopus database, 1 from the Google Scholar

database, 4 from the Science Direct database, 2 from

the ACM database and 3 from the IEEEXplore

database, remaining 369 papers.

Table 1: Quantity of work by repositories.

String

“machine learning algorithm” AND “education” AND

“games”

Repositories

Google Schoolar 78

Science Direct 169

ACM 13

IEEE 22

Scopus 100

TOTAL 382

After realizing the proposed exclusion criteria

were eliminated 328 works by the reading of the title

and the abstract, resulting in 41 articles for detailed

reading. The 41 works were read entirely because

they are more focused on the objective of this

research. However, from 41 articles, 36 articles were

eliminated because they applied classification

algorithms or were not directly related to games in

education, remaining 5 works. The 5 works were

A Systematic Mapping on Machine Learning Algorithms and Gamiﬁcation Applied to Education

355

selected because they applied machine learning

algorithms at each game level.

4 RESULTS

This section presents the results related to the answers

for the previous sections.

4.1 What Are the Games Styles, the

Area and the Level of Education

(Special or Not)?

The games that apply the machine learning

algorithms are quite diverse. Some of them were

developed in a specific way to attend the study, such

as MCP Quest and Skill Tree (Barata et al., 2016) in

the computer sciences area, and Romeo and Juliet

(Siu et al., 2018) in the languages area. Other works

used games available on the market such as Super

Monkey Ball 2 in the work of Cowley et al. (2014) in

the area of psychophysiology, games available on

Google Play in several areas such as sports, action,

strategy and Role-Playing Game (RPG) in the study

of Bharathi et al. (2016) and the Pacman game used

in Llorens-Largo et al. (2016) work. The last two

games mentioned were used in the area of computer

science. The computer science area concentrates

most of the application of machine learning

algorithms in educational games. Table 2 presents the

application areas of games with machine learning

algorithms.

Table 2: Games used in the areas and educational levels by

the authors.

Game Area Educational

level

Author

Super

Monkey

Ball 2

Psycho-

physiology

Higher

education

Cowley et al.,

(2014)

MCP

Quest and

Skill Tree

Computer

Science

Higher

education

Barata et al.

(2016)

Various

games

from

Google

Play

Computer

Science

Multiple

Levels

Bharathi et al.

(2016)

Pacman Computer

Science

Higher

education

Llorens-Largo

et al. (2016)

Romeo

and Juliet

Languages Multiple

Levels

Siu et al.

(2018)

Cowley et al. (2014) carried out in the area of

psychophysiology describes a new method to

integrate player preferences, experimental data and

game design patterns in a single framework, Play

Patterns and Xperience (PPAX). The framework

explored the patterns of gameplay and physiological

reactions of the players' faces, resulting in

information on reaction patters, moves and the

personality of each player.

In the work by Barata et al. (2016), the experiment

characterized data to predict the type of student at the

beginning of the master's course in computer science

at the University of Lisbon. For this, it used machine

learning algorithms to classify student data from one

period and predict the type of student in another

period. Through the inclusion of games in the

teaching grid, they created the games MCP Quest and

Skill Tree to expand interactivity and created a

ranking to reward the progression of levels and the

experiences acquired by students.

In the work of Bharathi et al. (2016), in the area

of computer science, several games and applications

installed on smartphones were analyzed to identify

and list design characteristics and gamification

elements such as challenges, feedbacks, rewards,

objectives, characters, badges, punctuation, levels,

ranking and dynamics of game states. These elements

of gamification can promote student motivation to

interact with the game more often and gain new

knowledge.

The work of Llorens-Largo et al. (2016) in the

area of computer science, developed the LudifyMe

system to assess the potential of gamification as a

means of improving learning. The main contribution

is in the application of games in the teaching of

artificial intelligence in order to improve motivation,

performance and student satisfaction.

In the work of Siu et al. (2018) in the area of

languages for learning the English language, they

developed a game platform based on the story of

Romeo and Juliet, which analyses the scores obtained

by the student to verify the progression and mastery

of the language. The purpose of the platform is to

predict whether the student has the capacity to

achieve an average score in the tests.

The systematic mapping verified and identified

that the educational levels of application are for

higher education, basic education and various levels

(basic to higher education), however, none of the

games was used for special education. However, there

is a predominance of works in higher education, and

analysing the amount of works at each level, there are

3 works related to higher education and 2 works that

fall into various levels (basic to higher education).

CSEDU 2021 - 13th International Conference on Computer Supported Education

356

4.2 How the Games Architecture Is

Built?

The games are divided according to type, for

example, serious games, racing, puzzle, strategy,

gamified web platform and RPG (Role Playing

Game). In the work of Barata et al. (2016) serious

games were developed containing stages and

missions. As the missions are carried out, a more

complex one is presented and the players receive

scores when they finish, however the missions can be

executed in several ways as long as all objectives are

fulfilled. In the game MCP Quest, for example, it is

an online treasure hunt in which students access a

page containing a multimedia artifact that must be

edited and manipulated to unravel the clue to access

the URL of the next page. As the missions are

completed the player earns experience points and is

able to access a more difficult mission.

Similarly, the game Skill Tree (Barata et al. 2016)

presents a tree of missions to be performed. At the

root of the tree, three initial missions are presented

and enabled and, in their branches, there are other

disabled missions. After completing the first

missions, the player receives the experience points

and new missions are enabled from the tree for

execution. If the player does not have enough

experience points, new missions of a more complex

level are not enabled. The architecture of this game is

server client, in which students access the game

through the pages provided and perform activities on

the platform. Thus, the data of all performed activities

are recorded in a database for later application of the

machine learning algorithm to understand the

students’ behavior.

The LudifyMe study by Llorens-Largo et al.

(2016) used the Prolog language to teach artificial

intelligence about the Pacman game. The game is

divided into difficulty levels according to the mazes,

amount and speed of movement of the main

character's enemies. As one level is overcome, the

next has the level of difficulty increased, and so on.

In this work, the labyrinth creation activities of the

game were carried out on a gamified web platform, in

which students must program the rules and relations

of the game and send for evaluation in a prediction

system. The architecture is composed by the gamified

platform, a database of events in which the

characteristics of the created labyrinths are extracted,

and an evaluation of the predictive system that

presents the results.

After the student submits the code for evaluation

on the platform, the information is sent to an event

database for further processing of the developed

labyrinths’ characteristics by the students. The

prediction system applies machine learning

algorithms to the database and presents the mazes’

results. With that, students and teachers can evaluate

the results and optimize the maze according to the

information provided in graphs.

Cowley et al. (2014) developed a framework,

called Play Patterns And eXperience (PPAX), to

analyze a set of data containing psychophysiological

information about the player’s facial expressions such

as tension, alertness, joy, upset, stress, calmness,

relaxation, etc., and video data containing plays that

were made.

The game used to apply the framework was Super

Monkey Ball 2, whose main mechanics is to roll a

ball, containing a monkey inside, on platforms of

different heights without dropping the ball and

reaching the end of the course. During the course the

player is rewarded with points for collecting bananas.

This framework analyses game event data and

psychophysiological data by applying the Frequent

Pattern Growth machine algorithm, which looks for

patterns in large data sets, to identify game design

patterns.

In the work of Cowley et al. (2014), Barata et al.

(2016), Bharathi et al. (2016), Llorens-Largo et al.

(2016), Siu et al. (2018), machine learning algorithms

are applied on a database formed by the data collected

during the interactions performed by the user in the

games. The information in the databases is treated to

facilitate the acquisition of knowledge through the

learning of machine algorithm with successive

training. And later, the information obtained is

analyzed and revised to improve the performance of

the machine learning algorithm. Finally, with this

information, evaluations are made to understand the

student's learning and the execution of their activities.

4.3 What Kind of Machine Learning

Algorithms and Techniques Are

Used in the Games?

Considering the selected works, it was identified that

the most used machine learning algorithms are those

that use supervised learning methods, such as the

Support Vector Machine (SVM) and the Naive Bayes

classifier. However, it was found that there is no

preference for a particular machine learning

algorithm, as they are chosen according to the authors'

preference and the context of each application.

Cowley et al. (2014) applied the Unsupervised

Frequent Pattern Growthe K-Means algorithms. In

the work by Barata et al. (2016) were used

Expectation Maximization Algorithm (EM), Naive

A Systematic Mapping on Machine Learning Algorithms and Gamiﬁcation Applied to Education

357

Bayes, Bayesian Networks, Sequential Minimal

Optimization (SMO), Logistic Regression K-Nearest

Neighbors (KNN).

Bharathi et al. (2016) applied the Support Vector

Machine (SVM), Naive Bayes, IBK, Decision Trees,

Random Forest, Sequential Minimal Optimization

(SMO) and J48 to rank games to a maximum score

and indicate whether they are successful or not.

Llorens-Largo et al. (2016) applied the Support

Vector Machine (SVM) to program Pacman game

rules and actions to interactively teach artificial

intelligence lessons to computer science students.

Finally, Siu et al. (2018) used the Decision Trees,

Support Vector Machine (SVM) and K-Nearest

Neighbor (KNN) algorithms. Table 3 presents the

machine learning algorithms used in their respective

years.

Table 3: Quantity of machine learning algorithms per year.

Algorithms Classification 2014 2016 2018

Frequent

Pattern Growth

Unsupervised 1

K-Means Unsupervised 1

EM Unsupervised 1

Naive Bayes Probabilistic 2

Bayesian

Networks

Probabilistic 1

Logistic

Regression

Supervised 1

K-Nearest

Neighbors

(KNN)

Supervised 2 1

Support Vector

Machine

(SVM)

Supervised 2 1

Decision Trees Supervised 3 1

Sequential

Minimal

Optimization

(SMO)

Supervised 1

TOTAL 2 1 5

The systematic mapping indicated that the

researched works between the years 2003 and 2020,

have a considerable variation in the use of machine

learning algorithms, however the repetition of

machine learning algorithms was identified, as shown

in Table 3. The most used algorithm is Decision

Trees, followed by Naive Bayes, K-Nearest

Neighbors (KNN) and Support Vector Machine

(SVM). In 2014, only 2 algorithms were used,

Frequent Pattern Growth and K-Means. In 2016, a

greater amount of work was applied, totalling 11

algorithms. In 2018, only 5 algorithms were used.

One observed factor was the use of supervised

learning algorithms in most of the studies, where the

data sets are already labelled for training containing a

predicted or desired response.

5 TRENDS

Machine learning played an important and significant

role in education because it allows exploring various

possibilities through which the system can perform a

cognitive analysis using a base with a set of input

data. Especially in an inverted classroom model, in

which the student-centered approach is adopted;

machine learning can be a revolutionary approach to

meeting student requirements based on their existing

skills (Naidu et al., 2018).

The systematic mapping process presented in this

article showed that there are few educational games

that are implemented with machine learning and

gamification. The use of machine learning allows

applying algorithms in a systematic way with the

ability to learn and adapt to changes in a system, using

criteria from examples of data or past experiences.

This learning ability not only predicts a certain

condition, but tries to find solutions to unpredictable

situations, as certain tasks or events that arise during

the execution of a system can be too complex to be

solved using classical programming (Alpaydin,

2014). The works analyzed in the systematic mapping

use machine learning after the execution of the game

and not at the execution level.

According to Samuel (1959) a scientist who

created the first learning program for checkers,

describes that machine learning transfers the ability to

the computer to learn without being programmed or

stated explicitly. In his experiment, every move made

by a person was analyzed and the machine learned

from mistakes and successes. Thus, each match

played by the computer gave the opportunity for

improvement, and in the next moves or matches,

enabled it to make tactical predictions of the game.

Therefore, any additional event and not foreseen in

the execution of a given move, made the system

interpret it as a probable error, store the information,

and then learn to avoid a possible repetition.

One of the characteristics of machine learning is

to generalize each event experienced, transform it

into a form of training and in subsequent executions

generate a hypothesis or predict by reaching a more

approximate estimate of a new process in question. It

also allows performance to be increased in instances

of non-visible data, providing greater accuracy for

future data. However, machine learning does not have

a well-defined or structured functionality to be

optimized. As error events arise, they serve as a filter

CSEDU 2021 - 13th International Conference on Computer Supported Education

358

for adjusting learning errors (Awad and Khanna,

2015).

The learning process is a major resource in

generalizing problems, as it is triggered according to

historical experiences. Historical experiences or logs

(activity records) are used to form the collections of

data which are later consulted to form a machine

learning domain model. Resources present in the K-

Nearest Neighbors (KNN) algorithm serve as a

support for classification indicators and discover

patterns, Naive Bayes is used for categorization based

on the frequency of information, Expectation

Maximization (EM) to characterize data not yet

observed and more appropriately point out what each

student does during their studies in order to optimize

learning (Barata et al., 2016).

Machine learning algorithms allow planning in

the development of educational games by verifying

the motivations, such as the Support Vector Machine

(SVM), which for each data entry seeks to analyse

and define patterns. With this assessment, developers

can design features to retain more attention from

players by encouraging them to overcome challenges

(Bharathi et al., 2016). Xu et al. (2016) claim that the

Random Forest algorithm points out which aspects

can be incorporated into interactive platforms with

different combinations between machine learning

models allowing it to increase the overall result.

Gamification can be used in the development of

educational games as a way of engaging students to

overcome proposed challenges such as missions and

more complex levels by intensifying interaction with

games. And from this, with the events of the

registered plays and matches it becomes possible to

form a data base for later application of machine

learning algorithms. With a data base having well-

defined attributes, the main contribution of the

application of machine learning algorithms is to favor

the obtainment of reaction and play patterns, player

and student profiles, behaviour and performance, and

student satisfaction. The gamification application

requires that ideal elements for the domain be

identified. According to Denden et al. (2018) the

elements are: points, levels, progress bar, feedback,

avatar, badges, leader identification (leader board)

and chat. Gamification theory has increased in the

area of education for the teaching of various

disciplines Lee and Doh (2012) and Domínguez et al.

(2013).

We observed that the games are applied at various

educational levels, however it is noted that their

predominance is at the higher level. In addition, it is

clear that the application of games aimed at the

education of intellectual disabilities is non-existent,

as in this systematic mapping on the application of

machine learning algorithm in educational games,

only one work related to deaf-mute people was found.

The application of this resource is still quite limited

to the area of computer science and scarce in other

areas such as health, sales, commerce, engineering,

administration, environmental, business and

industrial. According to Hidalgo et al. (2020)

concerns about unequal access to new technologies

have given rise to several socioeconomic questions,

especially gender, age, level, education, income and

habitat.

6 CONCLUSION

This paper presented a systematic mapping in order

to identify how educational games use machine

learning algorithms. The application of learning

algorithms is performed to extract information after

the consolidation of a database, that is, after the game

is over, the data are analyzed by the algorithm.

An identified matter was the use of supervised

learning algorithms in most of the studied, with a

considerable concentration in 2016. Another finding

was the use of gamification and gamification

elements such as those illustrated in the studies by

Barata et al. (2016) and Siu et al. (2018) like

educational games of their own authorship that use

dynamic elements such as narrative, progression and

restrictions, mechanical elements such as feedback

and challenges, gamification components such as

missions, points, classification (ranking) and

achievements. The other works used existing games

on the market that apply elements such as

progression, restrictions, challenges, points and

rewards (Cowley et al., 2014), (Bharathi et al., 2016)

and (Llorens-Largo et al., 2016). As for future works,

educational games with machine learning and

gamification can be developed allowing students to

improve their knowledge on a subject and the

machine algorithm can provide solutions for

unpredicted situations.

ACKNOWLEDGMENTS

This work is funded by the European Regional

Development Fund (ERDF) through the Regional

Operational Program North 2020, within the scope of

Project GreenHealth - Digital strategies in biological

assets to improve well-being and promote green

health, Norte-01-0145-FEDER-000042.

A Systematic Mapping on Machine Learning Algorithms and Gamiﬁcation Applied to Education

359

REFERENCES

Alanne, K., 2016. An overview of game-based learning in

building services engineering education. European

Journal of Engineering Education, 41(2):204–219.

Alhammad, M. M. and Moreno, A. M., 2018. Gamification

in software engineering education: A systematic

mapping. Journal of Systems and Software,141:131–

150.

Alpaydin, E. Introduction to Machine Learning. 2014. 3. ed.

Cambridge, MA: MIT Press, 2014.

Awad, M., Khanna, R., 2015. Machine learning. In:

Efficient Learning Machines: Theories, Concepts, and

Applications for Engineers and System Designers.

Berkeley, 2015. 1–18.

Barata, G., Gama, S., Jorge, J., and Goncalves, D., 2016.

Early Prediction of Student Profiles Based on

Performance and Gaming Preferences. IEEE Trans. on

Learning Tech., 9(3): 272–284.

Bharathi, A., Singh, A., Tucker, C., and Nembhard, H.,

2016. Knowledge discovery of game design features by

mining user-generated feedback. Computers in Human

Behavior, 60:361–371.

Bodnar, C. A., Anastasio, D., Enszer, J. A., and Burkey,

D. D., 2016. Engineers at Play: Games as Teaching

Tools for Undergraduate Engineering Students.

Engineering Education, 105(1):147–200.

Borges, R. P., Oliveira, P. R., Lima, R. G., and De Lima, R.

W., 2018. A systematic review of literature on

methodologies, practices, and tools for programming

teaching. IEEE Latin America Transactions, 16:1468–

1475.

Bozkurt, A., Durak, G., 2018. A systematic review of

gamification research: In pursuit of homo ludens,

International Journal Game-Based Learning (8) 15–33.

Brito, A. and Madeira, C., 2017. Metodologias gamificadas

para a educação: uma revisão sistemática. Anais do

XXVIII SBIE, 1:133.

Calderón, A., Trinidad, M., Ruiz, M., and O’Connor, R. V.,

2018. Teaching software processes and standards: a

review of serious games approaches. Communications

in Computer and Information Science, 918:154–166.

Cordero-Brito, S. and Mena, J., 2018. Gamification in the

social environment: A tool for motivation and

engagement. In ACM International Conference

Proceeding Series, 640–643.

Cowley, B., Kosunen, I., Lankoski, P., Kivikangas, J. M.,

Jarvela, S., Ekman, I., Kemppainen, J., and Ravaja, N.,

2014. Experience Assessment and Design in the

Analysis of Gameplay. Simulation and Gaming,

45(1):41–69.

De Avila dos Santos, J. and Luis Castro de Freitas, A., 2017.

Gamificação Aplicada a Educação: Um Mapeamento

Sistemático da Literatura. Renote, 15(1.

De Avila dos Santos, J. and Luis Castro de Freitas, A., 2017.

Gamificação Aplicada a Educação: Um Mapeamento

Sistemático da Literatura. Renote, 15.

De Sousa Borges, S., Durelli, V. H., Reis, H. M., and

Isotani, S., 2014. A systematic mapping on

gamification applied to education. In Proceedings of the

ACM Symposium on Applied Computing, SAC ’14,

pages 216–222, New York, NY, USA.

Denden, M., Tlili, A., Essalmi, F., Jemni, M., 2017.

Educational gamification based on personality, 2017

IEEE/ACS 14th International Conference on Computer

Systems and Applications (AICCSA) 1399–1405.

Dicheva, D., Dichev, C., Agre, G., and Angelova, G., 2015.

Gamification in education: A systematic mapping

study. Educational Technology and Society, 18(3):75–

88.

Gentry, S. V., Gauthier, A., Ehrstrom, B. L., Wortley, D.,

Lilienthal, A., Car, L. T., Dauwels-Okutsu, S.,

Nikolaou, C. K., Zary, N., Campbell, J., and Car, J.,

2019. Serious gaming and gamification education in

health professions: systematic review. Journal of

Medical Internet Research, 21.

Hidalgo, et al., A., 2020. The digital divide in light of

sustainable development: An approach through

advanced machine learning techniques. Technological

Forecasting and Social Change. 150 (1).

Inocencio, F. C., 2018. Using gamification in education: A

systematic literature review. In:International

Conference on Information Systems 2018, ICIS 2018.

Kamunya, S., Maina, E., and Oboko, R., 2019. A

Gamification Model for E-Learning Platforms. In 2019

IST-Africa Week Conference, IST-Africa, 1–9.

Kapp, K. M., 2012. Theories Behind Gamification of

Learning and Instruction. The Gamification of Learning

and Instruction: Game-Based Methods and Strategies

for Training and Education, 51–74.

Kitchenham, B. and Charters, S., 2007. Guidelines for

performing systematic literature reviews in software

engineering. School of Computer Science and

Mathematics, 1–65.

Lara, J. A., Aljawarneh, S., and Pamplona, S., 2019.

Special issue on the current trends in e-learning

assessment. Journal of Computing in Higher Education.

Lee, J. and Hammer, J., 2011. Gamification in education:

What, how, why bother? Academic Exchange

Quarterly, 15(2):146.

Llorens-Largo, F., Villagrá-Arnedo, C. J., Gallego-Durán,

F. J., Satorre-Cuerda, R.,Compañ-Rosique, P., and

Molina-Carmona, R., 2016. LudifyME: An Adaptive

Learning Model Based on Gamification. In Caballé, S.

and Clarisó Learning Data Analytics and Gamification,

R. B. T. F. A., editors, Formative Assessment, Learning

Data Analytics and Gamification: In ICT Education,

pages 245–269.

Lopes, R. P., Mesquita, C., Rio-Rama, M. de la C. del, &

Alvarez-Garcia, J. (2018). Collaborative Learning

Experiences for the Development of Higher-Order

Thinking. ESPACIOS, 39(17).

Mora, A., Riera, D., González, C., and Arnedo-Moreno, J.,

2017. Gamification: a systematic review of design

frameworks. Journal of Computing in Higher

Education, 29(3):516–548.

Morelock, J. R., 2013. Systematic literature review: An

exploration of gamification in the context of

engineering education. In IIE Annual Conference and

Expo 2013, 453–462.

CSEDU 2021 - 13th International Conference on Computer Supported Education

360

Naidu, V. R., Singh, B., Farei, K. A., Suqri, N. A., Al-

Masri, A. N., Al-Assaf, Y. Machine Learning for

Flipped Teaching in Higher Education: A Reflection.

In: Conference Sustainable Devel. and Social

Responsibility. v. 2. Switzerland: ASTI, 2018. p. 129-

132.

Ortiz-Rojas, M., Chiluiza, K., and Valcke, M., 2017.

Gamification and learning performance: A systematic

review of the literature. In Proceedings of the 11th

European Conference on Games Based Learning,

ECGBL 2017, 515–522.

Osatuyi, B., Osatuyi, T., De La Rosa, R., 2018. Systematic

review of gamification research in is education: A

multi-method approach, Commun. Assoc. Inf. Syst.

(42) 95–124.

Peixoto, M. M. and Silva, C., 2015. Requirements for

gamified educational software: A systematic literature

review [requisitos para softwares educacionais

gamificados: Uma revisão sistem ́atica de literatura]. In

CIBSE 2015 - XVIII Ibero-American Conference on

Software Engineering, 618–631.

Rodrigues L. F., Oliveira A., Rodrigues H., 2019. Main

gamification concepts: A systematic mapping study.

Heliyon, v. 5, n. 7, p. e01993.

Samuel, A. L., 1959. Some studies in machine learning

using the game of checkers. IBM Journal of Research

and Development, v. 44, p. 210–229.

Seaborn, K. and Fels, D. I., 2015. Gamification in theory

and action: A survey. International Journal of Human

Computer Studies, 74:14–31.

Siu, W. L., Lim, T. S., Chen, Y. R., Chen, Y. L., Jou, Y. A.,

and Chen, Y. C., 2018. Using an english language

education app to understand the english level of

students. In 2018 27th Wireless and Optical

Communication Conference, WOCC 2018, pages 1–3.

Souza, M. R. A., Veado, L., Moreira, R. T., Figueiredo, E.,

and Costa, H., 2018. A systematic mapping study on

game-related methods for software engineering

education. Information and Software Technology,

95:201–218.

Stevenson, M. P., Hartmeyer, R., and Bentsen, P., 2017.

Systematically reviewing the potential of concept

mapping technologies to promote self-regulated

learning in primary and secondary science education.

Educational Research Review, 21:1–16.

Subhash, S. and Cudney, E. A., 2018. Gamified learning in

higher education: A systematic review of the literature.

Computers in Human Behavior, 87:192–206.

Tenório, M. M., Reinaldo, F. A. F., Góis, L. A., Lopes, R.

P., and dos Santos Junior, G., 2018. Elements of

gamification in virtual learning environments: A

systematic review. Advances in Intelligent Systems and

Computing, 716:86–96.

Trinidad, M., Calderón, A., and Ruiz, M., 2018. A

systematic literature review on the gamification

monitoring phase: How spi standards can contribute to

gamification maturity. Communications in Computer

and Information Science, 918, 31–44.

Truong, H. M., 2016. Integrating learning styles and

adaptive e-learning system: Current developments,

problems and opportunities. Computers in Human

Behavior, 55:1185–1193.

A Systematic Mapping on Machine Learning Algorithms and Gamiﬁcation Applied to Education

361