The APOGEE Software Platform for Construction of Rich Maze

Video Games for Education

Boyan Bontchev

, Dessislava Vassileva

and Yavor Dankov

Faculty of Mathematics and Informatics, Sofia University St. Kl. Ohridski, J. Baurchier 5 Blv., Sofia 1164, Bulgaria

Scientific Research Department, Sofia University St. Kl. Ohridski, Dragan Tsankov 8 Blv., Sofia 1164, Bulgaria

Keywords: Video Games, Maze, Generation, Platform, Game-based Learning, APOGEE.

Abstract: Nowadays, the integration of serious video games into educational and training processes tends to be more

and more popular. The present paper outlines the software architecture of an innovative online platform for

an automatized construction of educational video games, which is going to allow non-IT professionals such

as teachers, pedagogues, and educationalists to design, automatically generate and personalize educational

video games based on a formal descriptive game model. The games represent rich educational video mazes

providing didactic multimedia content personalized upon various characteristics of the player. The

construction process includes three stages: game design, game validation, and game generation. The

integration of analytics tools into the platform will monitor all of the platform's data and processes hence will

facilitate the platform users to make more adaptive, effective, and efficient video maze games for education.

1 INTRODUCTION

With the rise of the modern computer era, digital

games became a highly popular media thanks to their

visual interactivity and ability to represent stories,

knowledge, virtual worlds, and interactive objects in

an appealing and spectacular way (Salen and

Zimmerman, 2004). Most digital games are designed

as 2D or 3D video games and can introduce both

tangible and intangible cultural artefacts by providing

high immersion and engagement (Bontchev, 2016)

through “an integrated form of fun and play” (Gee,

2003).

Besides video games for fun, there exist also other

games designed with purposes different than

entertainment. The term serious game was introduced

by Abt (1970), who defined such a game as having

“an explicit and carefully thought-out educational

purpose and are not intended to be played primarily

for amusement”. The majority of serious, or so called

applied, games are designed for educational or

training purposes, however, there are many such

games applied for defense, advertising, marketing,

https://orcid.org/0000-0002-8554-2188

https://orcid.org/0000-0002-3169-0097

https://orcid.org/0000-0003-3670-8599

http://apogee.online/index-en.html

political votes, industry control, and scientific

research (Sawyer and Smith, 2008).

Serious games serve as an engaging and

motivating means for game-based learning (GBL)

and training, however, their design and development

require a higher production cost (GALA, 2011). On

the other hand, there are few platforms for free of

charge, online creation of serious games (Bontchev

and Panayotova, 2017). Hence, educational and

training institutions cannot afford a variety of

educational video games dedicated to different

learning domains, that restricts the scale of modern

GBL to isolated cases in some schools and

Universities.

The present paper outlines the software

architecture of an online platform for an automatised

creation of educational video games being under

construction in the scope of the APOGEE (smArt

adaPtive videO GamEs for Education) research

project

. This innovative open platform is going to

allow non-IT professionals such as teachers,

pedagogues, and educationalists to construct,

automatically generate and personalize educational

Bontchev, B., Vassileva, D. and Dankov, Y.

The APOGEE Software Platform for Construction of Rich Maze Video Games for Education.

DOI: 10.5220/0007930404910498

In Proceedings of the 14th International Conference on Software Technologies (ICSOFT 2019), pages 491-498

ISBN: 978-989-758-379-7

491

video games based on a formal descriptive game

model including semantic structuring of both game

and didactic content. The games represent rich

educational video mazes defined as a 3D maze video

game providing rich didactic multimedia content

personalized upon various characteristics of the

player (i.e., the learner) and presented within the

maze halls not only on learning boards but as well as

within puzzle mini-games of various types embedded

into each hall. As well, a rich educational maze is

supposed to provide rich gaming and learning

experience thanks to including intelligent virtual

players and applying a dynamic, player-centric

adaptation of both difficulty of learning tasks and the

audio-visual properties of the game environment.

The paper is structured as follows: after the

introduction, we present our motivation to the

construct and maintain such an open software

platform for construction of rich maze video games

for education. We outline some of the major problems

with production of the serious games and, as well, the

existing tools and platforms for construction of

educational video games. Next, we present the

process of construction of rich video maze games for

education and the software architecture of the

platform. In the fourth section of the paper, we

discuss some of the services provided by the platform

and the micro-services design pattern for their

implementation. Finally, we conclude with some

remarks about the importance and the practical

application of the platform being under construction

and, as well, provide some directions about our future

works.

2 BACKGROUND

The section presents our motivation to create and

maintain the APOGEE software platform for the

construction of rich maze video games for education,

together with a short review of similar works in the

area of automatized construction of educational

mazes and puzzles.

2.1 Problems with Production of

Serious Video Games for Education

Modern GBL needs various cheap, attractive, and

engaging educational video games having a quality

similar to the entertainment games and able to embed

content from different learning domains. Nowadays,

these requirements to the serious games applied for

education appear not to be satisfied, at least of several

reasons:

• Serious games have a higher ratio between

development costs and the number of potential

players – educational games have a specific

purpose and are targeted to a very limited auditory

compared with the commercial games (GALA,

2011);

• Due to the very limited budget of the educational

and training institutions, serious games for

learning are rather less attractive than games for

fun;

• Construction of an educational game requires the

inclusion of teachers and pedagogues as core

game designers into the overall production

process, which is not the usual practice in game

studios (Paunova, 2019);

• There is a worldwide lack of free and

customizable platforms for the automatic creation

of educational games (Bontchev and Panayotova,

2017).

Besides the serious problems with serious games

listed over, the GALA Roadmap (2011) identifies

several other obstacles and open issues hampering the

massive penetration of GBL as a modern method of

technology-enhanced teaching and training:

• Difficult and non-adequate accordance between

the game mechanics and the learning paradigms;

• Lack of a reference framework for achieving

compliance of the gameplay issues with the

learning objectives;

• Due to a shortage of effective tracking of

individual learner progress, educational games

could hardly be applied for assessment

purposes;

• Educational games need smart and realistic virtual

players, i.e. NPCs – for this purpose, adequate

psychological theories should be applied

together with modern AI techniques.

Next to Gala, Shapiro (2014) outlines ten crucial

obstacles hampering video games to be applied in

education. He pointed out three important

technological handicaps preventing a wide usage of

educational games, namely purchasing cost,

problematic discovery of video games suitable for a

specific curriculum, and uncertain ways for applying

video games into the teachers' practice.

All these problems stated over made us start the

construction of an open and free software platform for

the construction of educational video games.

2.2 Similar Works

Undoubtedly, many people and organisations try to

answer problems with the production of serious

games for education. The need of platforms and tools

ICSOFT 2019 - 14th International Conference on Software Technologies

492

for an automatized construction of educational games

led to appearance of such software especially for

creation of mazes, quizzes, and puzzles as simple

single-player video games missing intricate narrative,

high interactivity, and complex character

development.

Up to present, there are available only a few and

simple online tools for automatized creation of mazes

and puzzles. Quandary

is a very popular tool

facilitating the creation of 2D online action mazes.

Such mazes represent multi-stage scenarios

composed by states dedicated to a concept and having

several possible choices/actions to proceed within the

maze. After selecting an action, the player moves to

the next state of the transition graph and explores its

specific scenario. Action mazes were successfully

applied for game-based learning in foreign languages

(Kiliçkaya, 2017) and for improving decision-making

professional skills (Gilbert and Priddle, 2010). Next

to Quandary, Qedoc Quiz Maker

proved to be

another popular freeware for creating and distributing

interactive educational and training modules. The

versatile playback environment of Qedoc Quiz Maker

can serve not only as a player of quizzes with

questions of hundred different types but as a system

for exam revisions, a learning tool including

generators of mathematical problems, or a flexible

manager of surveys. Recently, another maze

generator was proposed within the scope of the

ADAPIMES research project (Bontchev and

Vassileva, 2017). The generator created 3D adventure

mazes with puzzles for unlocking doors to next maze

rooms, whereupon teachers were able to customize

maze structure and add their preferred content to the

maze rooms. However, there are no examples of

customizable and personalizable mazes available

online neither of platforms for generation of such

mazes, which could be easily customized for various

educational curricula.

3 THE APOGEE PLATFORM

FOR RICH MAZE VIDEO

GAMES FOR EDUCATION

Based on several previous experiences in the area of

generation of personalized learning paths (Vassileva,

2012), emotionally-adaptive learning games

(Bontchev and Vassileva, 2017), and educational

video mazes (Bontchev and Panayotova, 2017), the

http://www.halfbakedsoftware.com/quandary.php

APOGEE project develops the idea of an automatic

generation of rich video maze games for education.

This section outlines the paradigms of rich

educational video maze games and explains their

construction process and the software architecture of

an online maze game platform planned to be

developed until 2020.

3.1 Rich Educational Maze Games

The APOGEE online platform will allow automatized

construction of rich video maze games for

educational purposes. Such games apply as a game

container a 3D planar video maze, whose halls

(rooms) represents rich didactic multimedia content

by means of:

• learning boards;

• puzzle mini-games of various types

representing learning tasks;

• smart virtual players, or non-player characters

(NPCs), providing help and answers to the

player questions.

The didactic content can be personalized upon

various characteristics of the game player/game-

based learner model such as:

• Demographic characteristics like age and

gender;

• Learner/player characteristics:

o Static parameters – goals and preferences,

knowledge level, learning-playing style;

o Dynamic features – effectiveness,

efficiency, and speed of solving tasks.

The personalized didactic content of the learning

tasks is presented by puzzle games embedded into

maze halls and having various types, such as

answering a question or quiz for unlocking a door,

arranging a pre-generated 2D puzzle, solving a ‘word

soup’ puzzle, rolling balls marked with both text and

texture to certain positions or objects, detection of

hidden objects and classifying them by specific

feature, and memory or shooting games. Solving

puzzles in a maze hall may be mandatory or optional.

The player should solve all the mandatory puzzles in

order to proceed to the next hall of the maze. Optional

puzzles might be solved just for increasing the

learner’s score, or just for fun.

All the characteristics of the player/learner model

serve for personalization of learning content and, as

well, for dynamic, player-centric adaptation of

difficulty of learning tasks (presented by puzzles), the

https://www.softpedia.com/get/Others/Home-Education/

Qedoc-Quiz-Maker.shtml

The APOGEE Software Platform for Construction of Rich Maze Video Games for Education

493

audio-visual properties of the game environment, and

the NPC behavior.

3.2 APOGEE Construction of Rich

Educational Maze Games

The construction of rich educational maze games

using the APOGEE platform includes three stages:

game design, game generation, and game validation.

3.2.1 Game Design

Rich educational maze games use a planar maze

consisting of halls connected each other by means of

doors. By default, the doors are locked and need to be

unlocked by the player by providing the correct

answer to the door question. After unlocking a door,

the player can open it by a mouse click on the door

and, next, proceed to the next hall. If a hall contains

mandatory puzzles, they should be solved in order to

answer the unlocking question. The learning content

of a hall of spread on learning boards (canvases),

puzzles, and an NPC available at the hall for helping

the learner. Besides learning content, a hall has also

gaming content consisting of various assets like 3D

gaming objects, music and sounds to be played in the

hall and at specific situations; illumination,

decoration, and textual help messages (parameterized

at design stage), and textures for the walls, floor, and

ceiling.

The game construction process is presented in fig.

1. Currently, game designers need to describe their

game formally within an XML document presenting

both the learning and gaming contents. As far as only

a third of the surveyed teachers are definitely positive

about constructing an XML design document for their

games (Bontchev and Panayotova, 2017), the project

team develops an online drag-and-drop maze editor

for facilitating the maze design. The editor is

controlled by the maze XML Schema (i.e., an XSD

document) in order to reflect future changes in the

organization of maze halls. It consists of:

• A connectivity editor serving for defining the

connections between the maze halls;

• A property editor facilitating the design of each

maze hall including content for the learning

boards, definition of embedded mini-games,

and all needed gaming assets together with

properties of the available NPC.

The didactic content for both the learning boards

and the puzzles could be defined in several versions

in order to be personalized at the beginning of the play

according to the static and dynamic characteristics of

the learner model (ADAPTIMES, 2019).

At each moment of the design process, the maze

designer can generate and download the XML

document defining formally the designed maze game.

The generated XML is a valid instance of the XML

Schema provided to the editor.

The XML document describing the maze game

should be validated by a XML Schema (XSD file)

before starting the maze game generation process.

Fig. 2 represents a scene of the roll-the-balls puzzle

game automatically generated by means of a XML

document describing a sample history maze

(Terzieva, 2019). There are shown two balls having

on their upper canvases blazons, and four rings

having appropriate titles. The goal of this puzzle

game is to roll each ball to its matching ring.

3.2.2 Game Generation

Among the many types of serious games, we have

chosen rich educational maze games because they can

be easily generated automatically by applying a

formal, XML description of the maze game, together

with all needed game assets. The maze generation is

possible by means of a custom plugin named Maze

Builder and developed for the Unity 3D game

platform (Unity, 2018). For an offline, local

generation, the plugin should be imported into the

Unity game editor as a custom package. Next, it

requires entering an XML document valid against the

XML Schema and describing a maze game, together

with an archive of all the gaming assets. Having the

XML maze definition document and the asset

archive, the Maze Builder plugin generates the maze

in few seconds. The generated maze can be viewed

and, if needed, updated in the Unity visual game

editor as shown in fig. 2. For example, the designer

could change the didactic content presented on a

learning board, or the position of a hidden object or

of a destination circle on the floor for rolling a ball to

it.

Finally, the maze designer is supposed to do a

build of the generated maze game for a given platform

such as PC desktop, Web browsers, or mobile

devices. For the future, the processes of generating

the maze and building the video game for a specific

platform are going to happen online, without any

additional intervention by the game designer.

3.2.3 Game Validation

After the automatic generation of a rich educational

maze and building the maze video game for a specific

platform, the game designer should validate the

constructed video game by playing it.

ICSOFT 2019 - 14th International Conference on Software Technologies

494

Figure 1: A view of the APOGEE maze games construction process (the modules in dotted line are under development).

For the moment, rich educational maze video

games can be played easily only at PC desktop

platforms or Web browsers due to the complexity of

the 3D maze interactions. In order to validate the

game generation process, the game designer

(supposed to be a non-ICT people such as teachers or

pedagogues) should play at least once the maze game.

He/she should check the appearance of the didactic

content on both the learning boards and puzzle mini-

games, the location of generated hidden objects, the

interactivity and all the issues concerning the

gameplay.

Figure 2: A view of a generated maze game with a roll-the-

balls puzzle mini-game.

When playing the game through the online game

platform, the player will enter the game with his/her

personal ID, i.e. with his/her player´s model. Thus,

the personal characteristics of this player will be

applied at the beginning of the play for content

personalization and dynamic adaptation of the

gameplay. Hence, the designer can inspect all the

gameplay issues of the generated rich educational

maze and, if needed, to update the game model and

the XML game definition document and to launch

again the generation and build process.

3.3 Software Architecture of the

APOGEE Game Platform

Nowadays, one of the most common approaches used

for the implementation of distributed applications is

the micro-services architecture. This architecture is

based on SOA and it is built from a small one or more

services that can be deployed independently of one

another (Fowler and Lewis, 2014). Each micro-

service can exist as a standalone application and it is

responsible for a performing a task, part of the overall

workflow. Usually, these services are connected to

each other over HTTP and communicate between

themselves through an interchange of messages

(Dragoni et al., 2018).

The APOGEE Software Platform for Construction of Rich Maze Video Games for Education

495

Figure 3: The APOGEE software platform architecture.

The micro-services approach assures scalability,

maintainability, easy integration, and

decentralization.

Precisely, because of the above-mentioned

advantages of the micro-services, the software

architecture of the APOGEE game platform follows

this architectural design pattern. As it is shown in fig.

3, the APOGEE architecture consists of a

presentation layer, several web services and a

persistent layer responsible for data storing. Each one

of the web services is implemented as a standalone

application and can be developed and changed

independently by others. The web services are

separated into three groups such as follows:

• Games construction services – it consists of six

web services (maze editor, maze validator,

maze creator, XML builder, game builder, and

asset manager) responsible for a game

construction. The maze editor allows a game

labyrinth to be created / edited and store its

structure in an XSD file and in the Game DB.

The maze validator validate the XSD file

produced by the maze editor;

• User management services – it includes four

web services (authentication, user profile,

learner profile, and player profile) related to

management of profiles of different users;

• Play game services – it contains three web

services (score viewer, player manager, and

learning and gaming analytics). The analytics

will provide the data for processing, analysing

and extracting valuable knowledge and

information from it.

The APOGEE platform is accessible for two type

of users – game creator and game player. As it is

shown in fig. 3, the first one (game creator) uses the

web services of game construction and user

management, and the second one (game player) uses

the web services of play games and user management.

Both types of users communicate with the web

services through a presentation layer presented by the

user interface of the APOGEE system. First, the user

is authenticated by username and password. Then,

ICSOFT 2019 - 14th International Conference on Software Technologies

496

depending on his / her role (game creator or game

player) continues to create / edit games or to play a

game.

4 DISCUSSION

The central objective of the APOGEE project is to

create a software platform for construction and

generation of smart adaptive 3D video maze games

consisting of a metadata-driven maze editor and a

Unity3D-based maze builder using an adaptation

control engine, an intelligent question and answering

(Q&A) agent, and declarative game description and

semantically structured virtual representation of

artefacts. It is very important non-IT professionals

such as teachers, pedagogues, and educationalists to

be able to use an open platform to construct,

automatically generate and personalize engaging

educational video games. Hence, the target user group

for the APOGEE platform will include not only IT

users but also people having no or limited knowledge

in programming and data science (Dankov and Birov,

2018) such as teachers or pedagogues. The platform

will provide the possibilities of creation of

educational games for three main groups of users:

• Users with experience in XML design - they

will create the game's design document from

scratch with any plain text editor and then

upload it to the server;

• Users with an initial experience in mark-up

languages – they will create the game design

document using the XML templates provided

by the portal and then upload it to the server;

• Users with no experience in mark-up languages

- they will create the game in the online editor

and then generate the XML game design

document to be used for generating the maze.

The APOGEE platform addressed all the existing

problems and challenges listed in Section 2.1. First,

game development costs will be practically zero

because users will only design the educational games

and, next, generate and build it automatically. The

generated maze games promise to be more attractive

than today’s serious games thanks to the inclusion of

various puzzles, game assets, and intelligent virtual

players. Teachers and pedagogues will act as core

game designers into the overall production process,

using a free and customizable platform for the

automatic creation of educational games without any

need of outsourcing to game studios. As well, smart

services will help designers to tailor the gameplay

issues with the learning objectives and specific

curriculums. Personalization of learning content and

dynamic adaptation of difficulty will incur greater

motivation, engagement and flow among the learners,

following a complicated learner model and the design

of the adaptation mechanisms (ADAPTIMES, 2019).

Finally, the analytics tools integrated into the

platform will provide effective tracking and

monitoring of individual learner progress. The

analytics services are planned in three directions:

• Learning analytics – such as efficiency,

effectiveness and time to learn through games

by the learner;

• Gaming analytics – like efficiency,

effectiveness and time of play by the player;

• Additional analytics – an opportunity of

integration of new analytics tools for the

purpose of the platform, such as Business

Visual Analytics (Dankov and Birov, 2018) for

monitoring overall data and statistics of the

platform, game creators, and players.

For realizing all the platform services, the

architecture of micro-services has been preferred over

the monolithic or layered approach for developing

distributed applications. Micro-services expose their

functions to other services or applications through an

API and can be deployed and scalable independently.

The architecture of the APOGEE platform follows

micro-services design pattern. Its code is split up into

three composite web services (Game construction,

User management, and Play games) built around

three business contexts and each of the services is

composed of several small services having a single

responsibility. Thereby, it is achieved flexibility,

scalability, independence, and maintainability.

5 CONCLUSIONS

The paper presented an innovative open platform for

an automatized creation of educational video maze

games being under construction in the scope of the

APOGEE research project. This platform includes a

drag-and-drop editor for creating a game and provides

methods for automatic generation of adaptive video

maze games. Hereby, it allows non-IT people easily

to design and creates educational maze games.

Moreover, it provides possibility of different

pedagogical strategies to be embedded in the

educational maze games that greatly facilitates the

game process development and reduce its production

cost.

Modern video games tend to include automated

conversational entities such as virtual players (i.e.,

NPCs) playing the role of personal assistants doing

tasks for the player, competitors or opponents

The APOGEE Software Platform for Construction of Rich Maze Video Games for Education

497

(Adams and Rollings, 2006). They are well accepted

by real players in the way people interact with chat

bots as a regular part of a chat room. Question-

answering was proposed first in role-playing games

but appears to be very important for any game having

NPCs. The APOGEE approach plans to apply

question answering, where possible answers to a

question in a given domain are ranked and

incorporated in large-scale goodness polarity

lexicons by means of a semi-supervised way. Smart

NPCs should provide adequate answers to the

player´s questions, especially to those belonging to

the game learning domain.

As future works, we plan to conduct practical

experiments with the APOGEE platform. The

experiments will include validation of the usability of

the platform by non-IT specialists (meeting their

game design requirements) as well as an assessment

of the adaptability and usefulness of virtual agents.

Hereby, that will make easier for teachers to apply

game based learning at schools and Universities. The

APOGEE platform provides an open solution that

allows zero development cost, easy maintaining of

educational and gaming content, and applying of

specific pedagogical strategy in a video maze game.

Moreover, with the APOGEE the game creators will

be able to apply a dynamic, player-centric adaptation

of both difficulty of learning tasks and the multimedia

game assets that is a key factor for an effective

learning process. This will facilitate the designers to

develop the platform and the creators to make more

adaptive, effective, and efficient educational games,

for various learning domains. The players will have

their own statistical metrics of success and failures, as

well as an opportunity for competitions between

players for achieving best results.

ACKNOWLEDGEMENTS

The research leading to these results has received

funding from the APOGEE project, funded by the

Bulgarian National Science Fund, Grant Agreement

No. DN12/7/2017.

REFERENCES

Abt, C.C. 1987. Serious Games. University Press of

America.

Adams, E., Rollings, A. 2006. Fundamentals of game

design. Prentice-Hall.

ADAPTIMES. 2019. Declarative model of educational

maze games with semantic structuring of game and

didactic content. Project Deliverable D2.1,

http://www.apogee.online/results-en.html, last viewed

on May 20, 2019.

Bontchev, B. 2016. Serious Games for and as Cultural

Heritage, Digital Presentation and Preservation of

Cultural and Scientific Heritage, Issue No. 5, 43-58.

Bontchev, B., Panayotova, R. 2017. Towards Automatic

Generation of Serious Maze Games for Education,

Serdica J. of Computing, Vol. 11, No 3, 249–278.

Bontchev, B. P., Vassileva, D. 2017. Affect-based

adaptation of an applied video game for educational

purposes. Interactive Technology and Smart Education,

Emerald, ISSN: 1741-5659, 14 (1), 31-49. DOI:

10.1108/ITSE-07-2016-0023.

Dankov, Y., Birov, D. 2018. General Architectural

Framework for Business Visual Analytics, Business

Modeling and Software Design, Lecture Notes in

Business Information Processing, Proceedings of 8th

International Symposium BMSD 2018, Vienna,

Austria, July 2–4, 2018, Editors: Shishkov, B., Springer

Int. Publ. AG, ISSN 18651348, 280-288.

Dragoni, N., Giallorenzo, S., Lafuente, A. L., Mazzara, M.,

Montesi, F., Mustafin, R., Safina, L. 2017

Microservices: yesterday, today, and tomorrow.

Present and Ulterior Software Engineering, Springer,

Cham, 195-216.

Fowler, M., Lewis, J. 2014 Microservices,.

http://martinfowler.com/articles/microservices.html.

GALA. 2011. GALA Roadmap. Deliverable D1.6, Ver. 2,

November.

Gee, J. P. 2003. What video games have to teach us about

learning and literacy, J. of Comp. Entertainment, 1, 20-

28.

Gilbert, A., Priddle, J. 2010. Using an action maze to

develop problem-solving skills in family law.

Networks, 13, 24-29.

Kiliçkaya, F. 2017. Infusing Action Mazes into Language

Assessment Class Using Quandary. Chapter in Balkan

Educational Studies – 2017, ed. H. Asutay, Trakya

University, 223-231.

Paunova, E. 2019. Didactic mini video games – students’

and teachers’ point of view, Proc. of ISEIC’2019,

March 20-22, Czech Republic.

Salen, K., Zimmerman, E. 2004. Rules of play: Game

design fundamentals, MIT Press.

Sawyer B., Smith, P. 2008. Serious games taxonomy.

Presentation given at Serious Games Summit.

Shapiro, J. 2014. Games in the Classroom: Overcoming the

Obstacles, https://ww2.kqed.org/ mindshift/2014/

09/12/games-in-the-classroom-overcoming-the-

obstacles/, last viewed on May 20, 2019.

Terzieva, V. 2019. Game-based teaching in history – case

study in Bulgarian schools, submitted to

EDULEARN’2019, 1-3 July, Palma de Mallorca, Spain.

Vassileva, D. 2012. Adaptive e-learning content design and

delivery based on learning styles and knowledge level.

Serdica Journal of Computing, 6(2), 207-252.

Unity. 2018. Unity User Manual. Version 2018.3,

https://docs.unity3d.com/Manual/index.html, last

viewed on May 20, 2019.

ICSOFT 2019 - 14th International Conference on Software Technologies

498