<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>What do serious games developers search online? A study of GameDev StackExchange</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Philippe Tamla</string-name>
          <email>philippe.tamla@studium.fernunitat-hagen.de</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Thilo Böhm</string-name>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Christian Nawroth</string-name>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Matthias Hemmje</string-name>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Michael Fuchs</string-name>
          <xref ref-type="aff" rid="aff2">2</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>FernUniversität in Hagen, Faculty of Multimedia and Computer Science</institution>
          ,
          <addr-line>Hagen</addr-line>
          ,
          <country country="DE">Germany</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>Forschungsinstitut für Telekommunikation und Kooperation e.V. (FTK)</institution>
          ,
          <addr-line>Dortmund</addr-line>
          ,
          <country country="DE">Germany</country>
        </aff>
        <aff id="aff2">
          <label>2</label>
          <institution>Wilhelm Büchner University of Applied Science</institution>
          ,
          <addr-line>Pfungstadt</addr-line>
          ,
          <country country="DE">Germany</country>
        </aff>
      </contrib-group>
      <fpage>131</fpage>
      <lpage>142</lpage>
      <abstract>
        <p>The interest in serious games (SG) has grown very fast in the last decade. Yet the research community rarely studies SG and the challenges SG developers are facing while creating games that should educate, train, and inform players. In this research, we use Latent Dirichlet Allocation (LDA) topic model to analyze a popular Q&amp;A site (Game Dev StackExchange) in order to understand the thoughts and needs of SG developers. We analyze the most active and popular topics about SG development. Our study reveals that SG developers' needs are very diverse, ranging widely from simple script code to simulate 3D learning environments, over best practices to design educational games for specific target groups, to features that can increase the player's motivation and learning outcomes.</p>
      </abstract>
      <kwd-group>
        <kwd>Serious Games Development</kwd>
        <kwd>Q&amp;A sites</kwd>
        <kwd>Latent Dirichlet Allocation</kwd>
        <kwd>Topic Models</kwd>
        <kwd>Software Search</kwd>
        <kwd>Software Reuse</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>Introduction</title>
      <p>
        Serious Games (SGs) are video games intended to educate the players instead of pure
entertainment
        <xref ref-type="bibr" rid="ref10">(David R. and Sandra L. 2005)</xref>
        . The interest in SGs and their
application in education and businesses areas has increased very fast in the last decade
        <xref ref-type="bibr" rid="ref32">(Vasudevamurt and Uskov 2015)</xref>
        . SGs provide an engaging and self-reinforcing
context in which the player is motivated and instructed towards non-game events or
processes, including business operations, training, marketing, well-being and
advertisement
        <xref ref-type="bibr" rid="ref10">(David R. and Sandra L. 2005)</xref>
        . Thus, it is not surprising that many global
companies (like Microsoft, IBM, SAP, Adobe, Cisco, Google etc.) have started integrating
SGs and gamification techniques in their main business functions, processes, and
activities
        <xref ref-type="bibr" rid="ref2">(Anderson and Rainie 2010)</xref>
        . However, the creation of effective SGs is a
difficult and complicated process which requires an appropriate balance between game
design and instructional design, including an in-depth understanding of learning,
cognition, emotion, and play
        <xref ref-type="bibr" rid="ref8">(Brent and Bill 2014)</xref>
        . But the efforts, time and costs of
serious games development can be reduced if existing software assets and other
pedagogical components available on the market can be effectively reused and integrated into
the development process
        <xref ref-type="bibr" rid="ref35">(Wim V et al. 2016)</xref>
        . However, this requires to understand
what SG developers search, and which factors can influence how they locate and
select software
        <xref ref-type="bibr" rid="ref14">(Hucka and Graham 2018)</xref>
        . Existing research has mainly studied the
needs of developers for traditional software engineering, ignoring the challenges of
SG developers. But, SG development is different from other traditional software
engineering disciplines because of its specific challenges related to the theoretical
understanding of learning, cognition, emotion and play
        <xref ref-type="bibr" rid="ref6">(Bellotti et al. 2010)</xref>
        . But, to date
very little is known about the development and maintenance challenges of SG
developers. Therefore, we conducted a study by analyzing discussion threads on Game
StackExchange (GDX) site1 with the aim of understanding search behaviours and
practices of (serious) games developers.
      </p>
      <p>We used LDA topic model to identify major topics and areas of interests in SG
developers searched for. We discover our topics by using SG terms and the most popular
tags used in SG related posts. We rank our topics based on the number of documents.
Finally, we analyzed the most popular and active topics by considering the number of
documents assigned to each topic and their views. We conduct our study by using the
following research questions: RQ1: What issues do serious games developers search
online? RQ2: What are the most popular topics about serious games development?
This paper is organized as follow: We review related works (section 2), highlight our
research questions and describe our methodology to detect SG-related posts (section
3). Then, we present and discuss our findings (section 4) and conclude (section 5).
2</p>
    </sec>
    <sec id="sec-2">
      <title>Related works</title>
      <p>The research most closely related to our work are those investigating SG development
and software search.</p>
      <p>
        Serious Games Development. Nowadays, we observe a growing number of
software engineers and researchers focusing on SG development
        <xref ref-type="bibr" rid="ref10 ref22 ref26 ref30 ref31 ref34 ref8">(Brent and Bill 2014;
David R. and Sandra L. 2005; Mestadi et al. 2018; Petridis et al. 2012; Söbke and Seicher
2016; van der Vegt et al. 2016)</xref>
        . Also, established business companies like Google,
IBM, SAP are motivated to design, develop, and implement full-scale SGs in their
business functions and processes because of the following advantages: SGs have the ability
to 1) solve complex problems collaboratively, 2) improve the efficiency of business
processes, 3) support predictive modelling and real-time visualization, 4) increase ROI2
from processes, time, and resources, and 5) provide more retention of knowledge
compared to traditional methods
        <xref ref-type="bibr" rid="ref32">(Vasudevamurt and Uskov 2015)</xref>
        . The creation of SGs is
also a complex process of game design, programming, content production, and testing
        <xref ref-type="bibr" rid="ref34 ref35">(Westera et al. 2016)</xref>
        . And its success significantly depends on the quality of external
technical gamification platforms, dedicated software architecture
        <xref ref-type="bibr" rid="ref35">(Wim V et al. 2016)</xref>
        ,
reusable SG engines, and advanced technology components (software assets) (Wim V
1 GDX is a popular Q&amp;A site used by thousands of (serious) game developers from
all around the world to post questions and discuss issues about game development.
2 ROI (Return on Investment) - https://www-01.ibm.com/software/rational/rsar/roi/
et al. 2016). However, there is a lack of standardization, best practices, and tools
        <xref ref-type="bibr" rid="ref32">(Vasudevamurt and Uskov 2015)</xref>
        for supporting the development of SGs and the reuse
of game assets which have the ability to preserve and enhance the games’ pedagogical
effectiveness
        <xref ref-type="bibr" rid="ref34 ref35">(Westera et al. 2016)</xref>
        . Thus, SG developers often rely on a large set of
entertainment-based features and game engines despite the inherent differences
between SGs and entertainment-based games
        <xref ref-type="bibr" rid="ref26 ref8">(Brent and Bill 2014; Petridis et al. 2012)</xref>
        .
But, the complexity, time, and effort for creating SGs can be reduced if existing tools
and game assets available on the market can be effectively reused
        <xref ref-type="bibr" rid="ref35">(Wim V et al. 2016)</xref>
        .
      </p>
      <p>
        Software Search is a common practice in Software Engineering that aims at
discovering suitable software for a given purpose
        <xref ref-type="bibr" rid="ref14">(Hucka and Graham 2018)</xref>
        . Existing
research has shown that Software Search is an important component of real, day-to-day
software engineering
        <xref ref-type="bibr" rid="ref29">(Singer et al. 2010)</xref>
        because it can increase work productivity (e.g.
through effective reuse). Studies investigating Software Search in the past have tried to
understand developers’ search practices and challenges
        <xref ref-type="bibr" rid="ref18">(Ko et al. 2006)</xref>
        including in
mobile development
        <xref ref-type="bibr" rid="ref27">(Rosen and Shihab 2016)</xref>
        . These studies relied on techniques, such
as, web survey, interview, search logs, or a combination of those
        <xref ref-type="bibr" rid="ref28">(Sadowski et al. 2015)</xref>
        ,
but the analysis of Q&amp;A discussion sites like StackExchange are becoming very
popular due to the increasing number of opensource tools and search engines available
online
        <xref ref-type="bibr" rid="ref27 ref5">(Barua et al. 2014; Rosen and Shihab 2016)</xref>
        . However, to the best of our
knowledge, there is no study that has studied SG development. Thus, in this research,
we will study what SG developers search and will examine their specific needs and
challenges while developing games for serious purposes.
3
      </p>
    </sec>
    <sec id="sec-3">
      <title>Research Methodology</title>
      <p>In this section, we describe our methodology to identify SG related posts. We present
and motivate our research questions. Then we describe our data collection and analysis
process which we use to answer our research questions.</p>
      <sec id="sec-3-1">
        <title>3.1 Research Questions and Motivation</title>
        <p>
          RQ1: What issues do serious games developers search online? – Past research on
Software Search has mainly studied what developers seek for general purpose software
engineering
          <xref ref-type="bibr" rid="ref14 ref29">(Hucka and Graham 2018; Singer et al. 2010)</xref>
          . But the analysis of SG
developers’ specific needs has remained almost ignored so far. Surprisingly, the literature
often argues that SG development differs from general purpose software engineering
          <xref ref-type="bibr" rid="ref25">(Murphy-Hill et al. 2014)</xref>
          because of the challenges related to the pedagogical
effectiveness of games
          <xref ref-type="bibr" rid="ref34 ref35">(Westera et al. 2016)</xref>
          which include the development of new
techniques and tools for training, educating, and motivating learners
          <xref ref-type="bibr" rid="ref10">(David R. and Sandra
L. 2005)</xref>
          . Therefore, we are motivated to analyze what SG developers search online and
what issues they may face.
        </p>
        <p>RQ2: What are the most popular topics about serious games development? –
Discovering the most popular SG discussions can help the SG community to identify major
areas of interest for future investigations. For instance, game companies may fine-tune
various aspects of their products and SG publishers may point to areas in which
practitioners have the most challenges and interest. Also, researchers may discover new
research opportunities with a high chance of having an impact on practice.
3.2</p>
      </sec>
      <sec id="sec-3-2">
        <title>Data Extraction and Data Processing</title>
        <p>
          To create our corpus, we first download the posts.xml file from the GameDev
StackExchange data dump (last updated June 2018) which contains user-generated
questions and related answers3. We parse the data dump which includes a total of
110,193 posts, from which 42639 are questions. For each extracted post, we maintain
a record of its metadata: the post type, timestamp and the total number of views. We
then process our corpus by removing all existing code snippets that are present in the
posts (i.e. enclosed in &lt;code&gt; HTML tags), as well as common English words which
do not help to create meaningful topics as demonstrated in past research
          <xref ref-type="bibr" rid="ref5">(Barua et al.
2014)</xref>
          . We then tokenized each word and remove all common English words (adverbs,
pronouns, and prepositions), and punctuations using the LingPipe library
          <xref ref-type="bibr" rid="ref4">(Baldwin and
Dayanidhi 2014)</xref>
          . We finally create our corpus by considering only question posts
because each question provides interesting insight about what the game developers have
tried. Answer posts oft discuss solutions to existing issues but do not highlight the
problem itself. Thus, we extract the title of each question, which nicely summarizes the issue
being discussed, and the related body which gives additional context information about
how this problem was initially approached by the asker.
3.3
        </p>
      </sec>
      <sec id="sec-3-3">
        <title>Identifying SG-related posts</title>
        <p>
          In order to found which posts of our corpus are relevant, we need a way to filter out posts which
are not really related to SG development. To achieve this goal, we examined prior work
          <xref ref-type="bibr" rid="ref5">(Barua
et al. 2014)</xref>
          and defined an initial set of SG-related keywords. We considered a list of SG
synonyms as proposed by (Michael Hoffmann 2016), a set of primary markets for SG
          <xref ref-type="bibr" rid="ref10">(David R. and
Sandra L. 2005)</xref>
          , and the top ten most popular SG engines
          <xref ref-type="bibr" rid="ref8">(Cowan and Bill 2014)</xref>
          as shown in
Table 1. Initial keywords used to find SG-related posts. However, with our initial keywords, we
couldn’t find enough posts that were suitable for our analysis. So, we applied the tag threshold
_ = to our collection algorithm (proposed by
          <xref ref-type="bibr" rid="ref27">(Rosen and Shihab 2016)</xref>
          ) to
.
identify more SG related posts by using the most popular tags assigned to each post previously
found with our initial keywords. Fig. 1 shows the top 30 most popular tags used in SG-related
discussions ranked by their popularity (the number of posts to which a tag is assigned). We
experimented different TRT_tag values and found that 35% was suitable for our analysis. With this
value, we were able to identify 18% more SG-related posts from our corpus. As shown in Table
2, our SG-related corpus contains in total 18902 posts. We believe that the list in our corpus
contains the most popular posts about serious games which we define here as SG-related posts.
        </p>
        <sec id="sec-3-3-1">
          <title>SG Synonyms</title>
        </sec>
        <sec id="sec-3-3-2">
          <title>SG Markets</title>
          <p>Applied Game – Applied Gaming – Serious Game – Serious
Gaming – Gamification – Games for Change – Games for Good
Military – Government - Education – Corporate-Games –
Healthcare – Politics – Religion – Art
3 To be found at: https://archive.org/download/stackexchange</p>
        </sec>
        <sec id="sec-3-3-3">
          <title>SG Engines</title>
          <p>218
152</p>
          <p>Unity – Flash – Second life – E-adventure – Unreal – Torque3d
– Cocos – Xna</p>
        </sec>
        <sec id="sec-3-3-4">
          <title>All posts</title>
          <p>
            SG posts found with initial keywords
SG posts found with TRT&gt;=35%
All tags assigned to SG posts
Tags relevant to SG posts
To discover the most popular SG topics, two metrics were introduced. First, the topic
relevance metric = represents the distance of a topic distribution
as!
signed by LDA to a document to the highest distribution assigned to that same
document. We use this metric as a threshold to consider only topics with the highest
distribution values as the most relevant ones. Usually, a document will have 1 to 5 most
dominant topics assigned to it with a distribution of 0,10 or higher
            <xref ref-type="bibr" rid="ref5">(Barua et al. 2014)</xref>
            .
However, due to the probabilistic nature of LDA, we found some topics with a very
small document distribution (e.g. 0.01) which introduced noises in our analysis. By
using an R(t) of 35 as our distribution cutoff, we were able to eliminate existing noises
in our analysis and could detect only relevant topics to our documents. Second, the
average number of views Av t = ,%-&amp;., '%'&amp;('%)&amp;*()(+ was introduced to analyze the
popularity of each topic t. TotalTopicDocuments is the number of documents assigned
to a topic t and SumDocumentsViews is the sum of views of all topic documents.
Existing research has shown that most of the online discussions are visited by passing (not
registered) users
            <xref ref-type="bibr" rid="ref20">(Mamykina et al. 2011)</xref>
            and that the number of views assigned to each
post may indicate the importance of a topic
            <xref ref-type="bibr" rid="ref27">(Rosen and Shihab 2016)</xref>
            . By using the
above-described measurements, each topic can be analyzed using two different
dimensions: the number of documents assigned to a topic indicating how active a topic is, and
the number of views of topic documents indicating the popularity of each topic in the
SG community.
3.5
          </p>
        </sec>
      </sec>
      <sec id="sec-3-4">
        <title>Topic Modeling using LDA</title>
        <p>
          Topic Modeling is a document clustering technique that uses statistical methods for
identifying abstract topics within a document collection. Several topic modelling
techniques have been developed and evaluated (such as Probabilistic Latent Semantic
Indexing, Non-negative Matrix Factorization)
          <xref ref-type="bibr" rid="ref11">(Ding et al. 2008)</xref>
          . In this research, we use
Latent Dirichlet Allocation (LDA) topic model as it was routinely applied to millions
of software engineering web documents
          <xref ref-type="bibr" rid="ref27 ref5">(Barua et al. 2014; Rosen and Shihab 2016)</xref>
          .
LDA uses the frequencies of words and topics (groups of words) from documents of a
text corpus to discover patterns and build a model based on related words and topics
          <xref ref-type="bibr" rid="ref27">(Rosen and Shihab 2016)</xref>
          . The number of topics (K) is defined by the user to control
the granularity of discovered topics and there is no standard value of K that can fit every
situation and dataset
          <xref ref-type="bibr" rid="ref13 ref33">(Grant and Cordy 2010; Wallach et al. 2009)</xref>
          . We experimented
different K values and found out that 30 yields good results without being too
restrictive. We also optimized our Dirichlet hyperparameters by using α = 0.5 and β=0.1 to
control how much fuzziness is allowed in the topics’ distributions across words and
topics respectively. It was shown that optimizing these the facto standard heuristics
parameters can dramatically improve the consistency in topic usage as the number of
topics increase
          <xref ref-type="bibr" rid="ref27">(Rosen and Shihab 2016)</xref>
          . To better classify the issues found in our
corpus, we add their dominant topic as determined by the LDA topic model. We
labelled our topics manually by considering the top words found in each topic and by
examining a sample of posts assigned to each topic.
4
        </p>
      </sec>
    </sec>
    <sec id="sec-4">
      <title>Results</title>
      <p>In this section, we present the results of applying our research methodology to answer
our research questions. We first examine the percentage of SG-related posts to all
existing posts in our corpus. Then, we describe each discovered topic and finally discuss
each topic qualitatively.
4.1</p>
      <sec id="sec-4-1">
        <title>Exploring the importance of serious games related posts</title>
        <p>Before delving into our research questions, we would like to understand the importance
of SG-related posts in our corpus. Thus, after extracting and processing our corpus, we
can now examine (as shown in Fig. 2) the number of SG-related posts (red line) and
their percentage (blue line) to all existing posts per month. Our analysis shows that at
least 36% of all existing posts are related to SG development. There is an increase of
SG-related posts from August 2010 to March 2013 reaching a maximal value of 300
2014 which finally gets stabilized over the following years till 2018 with an average of
27%. Thus, we conclude that the number of SG-related posts is growing and that
discussions about SG development are relevant in the game development community.
l
u
J
v
o
N
r
ä
l
u
J
v
o</p>
        <p>N
2010
2011
r
ä
l
u</p>
        <p>J
2012
v
o
N
r
ä
l
u</p>
        <p>J
2013
v
o
N
r
ä
l
u</p>
        <p>J
2014
v
o
N
r
ä
l
u</p>
        <p>J
2015
v
o
N
r
ä
l
u</p>
        <p>J
2016
v
o
N
r
ä
l
u
J
v
o
N
r
ä
2017
2018
M</p>
        <p>M</p>
        <p>M</p>
        <p>M</p>
        <p>M</p>
        <p>M</p>
        <p>M</p>
        <p>M
Percentage of SG posts</p>
        <p>SG posts per month</p>
      </sec>
      <sec id="sec-4-2">
        <title>RQ1: What issues do serious games developers search online?</title>
        <p>
          we merged some topics because of their semantical similarity. For instance, the topics
Collision Detection and Pathfinding are related to Game Physics and thus were merged
together. We arranged these topics in descending order according to their share metric
as proposed by
          <xref ref-type="bibr" rid="ref5">(Barua et al. 2014)</xref>
          . as proposed in
          <xref ref-type="bibr" rid="ref27">(Rosen and Shihab 2016)</xref>
          . In general,
we found that SG posts fall into one of the following topics: Game Design &amp; Learning
Design, (3D)Modeling/Rendering, Physics, Scripting, and Networking.
        </p>
        <p>Topic
Graphics Engine
Game Physics/Collision
Detection/Path Finding
Networking
Mesh
File&amp;Format
Scripting
Data structure
Platform
Game Design
Component
Learning Design
Camera
3D Modeling/
Rendering
Movement
Animation
Game Engine</p>
        <p>Top keywords
game java graphics creating render lwjgl library don function simple math framework multiple engines
question opengl threads testing people web
movement physics boxd speed libgdx simulation body path finding velocity projectile bullet gravity distance
ball collision detection network
server multiplayer game client side players network mmo online player networking architecture rts sync
prediction browser udp prevent communication clients
unity mesh rendering order error performance problem device correct triangle water android dynamic problems
udk collider vertices triangles text gui
xna file project monogame custom files content culling frustum format screen loading static write running
pipeline possible read asset creation
unity working script gameobject add properly audio object playing method new level doesn issue editor scaling
set child gameobjects isn
game data code structure objects implement build libgdx visual studio board designing font node cocosd rts
platforms ways gamemaker style
time canvas game html real games play javascript app google mobile step mode delta devices application run
simple services program
game games design good pattern idea creating practice making bad management gameplay effects mobile
common adventure patterns story book examples
system based component entity turn components systems entities particle architecture event ecs handling
design strategy combat battle oriented events types
learning design games learning api starting programming start platform card resources indie software
programmer computer developer started language development
camera screen position object mouse move moving libgdx set target issue coordinate person click shooter
transform touch follow environment full
multiple model models blender textures mesh single sprites objects parts unity rendering meshes building
support fbx separate techniques dealing combine
object point vector rotation direction rotate camera sphere rotating move calculate specific calculating plane
axis angle points quaternion find tower
animation character sprite unity create unityd animations skeletal states behavior flash jump controller bone
animate max android trees start animated
game engine open code source legal issues based engines design racing window assets browser small music
develop console put projects
To get a better insight about our discovered topics, we show a subset of examples
related to each group of topics along with some context information (like title, body, tags,
#postId) whenever needed, and will discuss their impact on SG</p>
        <sec id="sec-4-2-1">
          <title>Development.</title>
          <p>Game Design. We found one topic about game design. Our study reveals that SG
developers seek best practices and experts' opinions for taking important design
decisions, like choosing a proper design pattern (#32093) or selecting an appropriate
platform to create educational games (#8359).</p>
          <p>Title: Design Pattern for Social Game Mission Mechanics Body:
When we want to design a mission sub-system like in the The Ville
or Sims Social, what kind of design pattern / idea would fit the best?
There may be relation between missions (first do this then this etc...)
or not. What do you think sims social or the ville or any other
social games is using for this? I'm looking for a best-practise method
to contruct a mission framework for tha game. How the well-known
game do this stuff for large scale social facebook games? [ Topic:
Game Design, Tags: game-design,flash, social. PostId: #32093 ]</p>
          <p>Title: What makes a good educational game? – Body: I'm currently
creating a game framework/engine for educational games. My hope
is that this can be used in elementary schools. What makes a good
educational game? Which platforms should I target?
[ Topic: Game Design, Tags: serious-games cross-platform
educational-games.. PostId: #8359 ]
Also, SG developers seek techniques to implement AI for tracking the progression of
the player (#129492), assess the player’s learning outcomes (#94079) and interest
(#8925), and increase his engagement in learning (#149096, #28540).</p>
          <p>Title: Machine learning to improve strategy game AI - Body:I am
currently working on a simple strategy game as a hobby and I am
starting to think about designing an AI to add opponents ….My idea
was to record the player's activity every X amount of time as well
as the current state of some game variables. Then I would use some
machine learning algo (as a black box) to find a relationship
between game variables and actions to make. I would also record the
final score at the end of each game so that the AI learns more from
successful games than unsuccessful ones (or even better set a
score threshold above which to send the data to the AI)…
Any thoughts or directions towards interesting resources /tutorials
would be very welcome [ Topic: Level Design, Tags: game-design
ai strategy reinforcement-learning, PostId: #129492 ]
Title: How can I ensure the player learns all skills in a single open
level? Body: I am making a game that only has a single level. It is a
survival game. … I can have the player learn new skills gradually.</p>
          <p>For example, level one requires jumping, level 2 requires climbing,
level 3 requires both. Is there a common strategy to force the player
to learn all the necessary skills to succeed?</p>
          <p>
            Currently I have 4 directions which the player can travel from the
starting point. Each of which…[ Topic: Learning Design, Tags:
level-design tutorials reinforcement-learning, PostId: #94079 ]
Title: Things that make a game more interesting for kids – Body:
What makes a game more interesting or fun to play… Does the age
of consumer play an important role in his/her interest of game?What
can I put in a game to make it more fun, educational, and interesting
to play? – Topic: Game Design PostId: #8925 Tags: game-design
serious games
Title: How to calculate players engagement?- Body: What metrics
could be used to know what certain people think about the game
and how much they liked it? [ Topic: Game Design, PostId:
#149096 Tags: game-design -design metrics ]
Title:What are the downsides of only explicitly informing the player
of *success*? – Body: I working on educational software and
trying to increase engagement (“fun”) [ Topic: Game Design, Tags:
game-design, PostId: #28540 ]
There are two emerging research fields that focus on tracking and analyzing players’
interactions. First, Game Analytics (GA)
            <xref ref-type="bibr" rid="ref1 ref30">(A. All et al. 2016)</xref>
            focuses on developing
techniques to automatically follow users’ interactions and increase their engagement
(maximizing the time a player stays in the “flow”). Second, Learning Analytics (LA)
            <xref ref-type="bibr" rid="ref19">(Loh et al. 2015)</xref>
            aims at analyzing and measuring players’ learning outcomes, for
purpose of understanding and optimizing an environment in which it occurs. But, there is
a lack of standard formats to represent players’ interactions
            <xref ref-type="bibr" rid="ref9">(Cooper 2013)</xref>
            and facilitate
the processing and analysis of their learning progress (Ángel et al. 2017).
3D
          </p>
          <p>Modeling/Rendering. Our study reveals that SG developers seek instructions
and efficient tools or programs for creating and rendering (nice-looking) 3D objects
even with limited knowledge, skills, and time (#121634, #44168).</p>
          <p>
            Title: Creating a large amount of sprites and animations. Body: For
my AP U.S. Government and Politics class, I am making a spinoff
of Mortal Kombat. Because of time constraints, I'm using an open
source version of the game and want to create new characters. I have
about two weeks to do the project.. I want it to look as nice… What
Title: Software rendering 3d triangles in the proper order
Body: I'm implementing a basic 3d rendering engine in
software (for education purposes, please don't mention to use an
API). When I project a triangle from 3d to 2d coordinates, I draw
the triangle. If I'm sorting all the objects, this is n*log(n). Is this
programs or techniques would be best…? – [ Topic: 3D Model- the most efficient way to do this? [ Topic: 3D
Modeling/Rendering/Rendering, Tags: sprites animation, PostId: #121634 ] ing, Tags: 3D, rendering, software-rendering PostId: #44168 ]
3D Modeling is the process of using software to generate a representation of 3D
characters and objects for use in a game
            <xref ref-type="bibr" rid="ref12 ref16">(Dorribo-Camba et al. 2013; Jackson and Buxton
2007)</xref>
            , and Rendering refers to the process of creating an image from 2D or 3D model
            <xref ref-type="bibr" rid="ref21">(McShaffrey 2005)</xref>
            . Rendering covers every aspect ranging from texture creation,
lighting and shadow to materials, scene graph management
            <xref ref-type="bibr" rid="ref24">(Millington and Funge 2009)</xref>
            .
Well-designed SGs that can engage and affect learners require vivid characters, realistic
models and environment to be able to demonstrate behaviours with specific learning
lesson(s)
            <xref ref-type="bibr" rid="ref17">(Kelly et al. 2007)</xref>
            . Although existing research has revealed the reuse of
existing 3D models can reduce design time by 30%
            <xref ref-type="bibr" rid="ref12 ref16">(Dorribo-Camba et al. 2013; Jackson
and Buxton 2007)</xref>
            , there is still little work in the reuse of existing 3D models. Existing
techniques rely on indexing mechanisms to extract metadata (semi-automatically) from
models and support reuse
            <xref ref-type="bibr" rid="ref15">(Ioan Marius Bilasco et al. 2006)</xref>
            , however, substantial efforts
are still required to maintain the extracted metadata.
          </p>
          <p>
            Physics. Game physics (known as Physics) refers to the process of introducing
classical laws of mechanics in games
            <xref ref-type="bibr" rid="ref23">(Millington 2007)</xref>
            . Physics introduces the realistic
feel to the game and can stimulate real world objects (e.g. lighting, collision detection,
water, cloth physics, weather, etc.)
            <xref ref-type="bibr" rid="ref32">(Vasudevamurt and Uskov 2015)</xref>
            . Physics engines
support the modelling of virtual world objects by means of configuration to avoid
manual programming
            <xref ref-type="bibr" rid="ref30">(Söbke and Seicher 2016)</xref>
            . But physics calculations are often
computationally demanding and thus need to be appropriately integrated into game design
concepts in order not to distract the player, but promote immersive and engaging games
            <xref ref-type="bibr" rid="ref30">(Söbke and Seicher 2016)</xref>
            . We found that SG developers seek help to create and/or
optimize their algorithms for collision detection (#75109) or pathfinding (#108338),
and how to generate special characters (like living beings) and to integrate them in
specific virtual environments (like world politics) (#53137).
          </p>
          <p>
            Title: How to simulate early politics? – Body: I'm making a historically accurate game where the player can interact with past times and
shape the future. The entire game is scientifically generated with math and real physics…The player can build anything they want and it
will use physics to simulate each creation based on the materials it's made of. I was considering just modeling our political system and
then applying mathematical regression to the algorithm to regress it 100 million years. However, I'm worried about the limitations of
floating point numbers and I think that errors will compound.…P.S. I'm mostly talking about American dinosaurs, because they had the
most impact on world politics. [ Topic: Physics, Tags: mathematics procedural-generation simulations, PostId: #53137 ]
Existing game engines provide a variety of features and algorithms for AI, rendering,
and simulation that can be used right away by developers
            <xref ref-type="bibr" rid="ref30">(Söbke and Seicher 2016)</xref>
            .
Facilitating the detection of such features can lead to efficient SG development because
developers will only focus on developing domain-specific features instead of writing
low-level algorithms from scratch
            <xref ref-type="bibr" rid="ref10">(David R. and Sandra L. 2005)</xref>
            .
          </p>
          <p>
            Scripting. Our study reveals that SG developers seek specific code (like
unityscript) to control the behaviour of their games and do animation (#161824, #163929).
The literature refers to three levels of coding
            <xref ref-type="bibr" rid="ref7">(Blow 2004)</xref>
            : script code (for controlling
the overall content, rules, and high-level behaviour of games), gameplay code (for
graphics simulation, physics), engine code representing special algorithms for 2D/3D
animation, rendering, and programming networking. Coding plays a significant role in
the development of SGs and should be completely integrated into existing SG engines
in order to optimize the game’s efficiency in terms of performance and memory usage
            <xref ref-type="bibr" rid="ref32">(Vasudevamurt and Uskov 2015)</xref>
            .
          </p>
          <p>Scripting
File&amp;Format
3D Modeling
Level Design
Class/Object</p>
          <p>Platform
Game Physics</p>
          <p>Animation
Game Design</p>
          <p>Rendering
Average Views
NumberOfDocs</p>
          <p>
            Networking. We found that SG developers seek help to create games that allow
multiple users to play together over greater distance and on a distributed environment
like the Web (#157707). Networking technologies are responsible for the scalability
and security of games and also deal with issues related to bandwidth and latency
            <xref ref-type="bibr" rid="ref32">(Vasudevamurt and Uskov 2015)</xref>
            .
          </p>
          <p>Title: Optimization for end-to-end multiplayer pong-like game – Body: I've developing a small pong-like game for educational purposes
and I've reached the point where I'm looking at implementing network gameplay. I was looking into what optimizations are usually performed
for a smooth experience since latency can cause jitterry movements if not handled properly….</p>
          <p>I know that client-side prediction is used and read up on it but it seems to me that this is applies in client-server types of multiplayer. Is this
true, and if yes why?</p>
          <p>Topic: Networking, Tags: networking, multplayer, optimization, pong PostId: #157707
4.3</p>
        </sec>
      </sec>
      <sec id="sec-4-3">
        <title>RQ2: What are the most popular topics about SG development?</title>
        <p>0
2000
4000
6000</p>
      </sec>
    </sec>
    <sec id="sec-5">
      <title>5 Conclusion &amp; Future Work</title>
      <p>In this paper, we have used the LDA topic model on GameDev StackExchange corpus
to quantitatively detect and classify the main topics SG developers usually search. For
instance, we have found out that Rendering and Game Design are very active topics
(suggesting that SG developers are seeking best practices and common strategies to do
one thing), while discussions about Physics and Animation belong to the most popular
topics, which suggests that they are seeking help to fix their issues. Our study also
reveals that SG developers seek different things, ranging from simple scripts for
stimulating 3D learning environments, over best practices to design educational games, to
specific algorithms for increasing the player’s motivation and learning outcomes.
We believe that future research in SG should focus more on standardizing common
practices (like design patterns, design concepts) and tools to facilitate the development
of specific game genres and the reuse of common features that add a good balance of
learning and pedagogical objectives to games. This will motivate developers and
gaming companies to easily and efficiently design, develop, and deliver full-scale SGs while
being able to reduce the time to market. For this to be successful, a good understanding
of SG filed and theories (game design, learning theories, and domain content) is needed.</p>
    </sec>
    <sec id="sec-6">
      <title>Acknowledgement</title>
      <p>This research was partially funded by the European Union’s Horizon 2020 research and
innovation programme under grant agreement No 644187, the RAGE project
(www.rageproject.eu).</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          1.
          <string-name>
            <given-names>A.</given-names>
            <surname>All</surname>
          </string-name>
          ,
          <string-name>
            <given-names>E. P.</given-names>
            <surname>Nunez Castellbar</surname>
          </string-name>
          ,
          <string-name>
            <surname>J. Van Looy</surname>
          </string-name>
          (
          <year>2016</year>
          )
          <article-title>Game analytics maximizing the value of player data</article-title>
          ,
          <fpage>92</fpage>
          -
          <lpage>93</lpage>
          . Springer
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          2.
          <string-name>
            <surname>Anderson</surname>
            <given-names>JQ</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Rainie</surname>
            <given-names>H</given-names>
          </string-name>
          (
          <year>2010</year>
          )
          <article-title>Gamification: Experts expect'game layers' to expand in the future, with positive and negative results</article-title>
          .
          <source>Pew Internet &amp; American</source>
          Life Project
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          3.
          <string-name>
            <surname>Ángel</surname>
            <given-names>S-L</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Iván</surname>
            <given-names>M-O</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Jason</surname>
            <given-names>H</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Damon</surname>
            <given-names>R</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Andy</surname>
            <given-names>J</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Baltasar</surname>
          </string-name>
          (
          <year>2017</year>
          )
          <article-title>Applying standards to systematize learning analytics in serious games</article-title>
          .
          <source>Computer Standards &amp; Interfaces</source>
          <volume>50</volume>
          :
          <fpage>116</fpage>
          -
          <lpage>123</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          4.
          <string-name>
            <surname>Baldwin</surname>
            <given-names>B</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Dayanidhi</surname>
            <given-names>K</given-names>
          </string-name>
          (
          <year>2014</year>
          )
          <article-title>Natural language processing with Java and LingPipe Cookbook</article-title>
          . Packt Publishing Ltd
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          5.
          <string-name>
            <surname>Barua</surname>
            <given-names>A</given-names>
          </string-name>
          , W. Thomas S,
          <string-name>
            <surname>E. Hassan A</surname>
          </string-name>
          (
          <year>2014</year>
          )
          <article-title>What are developers talking about? An analysis of topics and trends in</article-title>
          <source>Stack Overflow</source>
          <volume>19</volume>
          :
          <fpage>619</fpage>
          -
          <lpage>654</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          6.
          <string-name>
            <surname>Bellotti</surname>
            <given-names>F</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Berta</surname>
            <given-names>R</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Gloria</surname>
            <given-names>A</given-names>
          </string-name>
          de (
          <year>2010</year>
          )
          <article-title>Designing effective serious games: opportunities and challenges for research</article-title>
          .
          <source>International Journal of Emerging Technologies in Learning (iJET) 5</source>
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          7.
          <string-name>
            <surname>Blow</surname>
            <given-names>J</given-names>
          </string-name>
          (
          <year>2004</year>
          )
          <article-title>Game development: Harder than you think</article-title>
          .
          <source>Queue 1</source>
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          8.
          <string-name>
            <surname>Brent</surname>
            <given-names>C</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Bill</surname>
            <given-names>K</given-names>
          </string-name>
          (
          <year>2014</year>
          )
          <article-title>A Survey of Frameworks and Game Engines for Serious Game Development</article-title>
          .
          <source>2014 IEEE 14th International Conference on Advanced Learning Technologies</source>
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          9.
          <string-name>
            <surname>Cooper</surname>
            <given-names>AR</given-names>
          </string-name>
          (
          <year>2013</year>
          )
          <article-title>Learning Analytics Interoperability - a survey of current literature and candidate standards</article-title>
          . http://blogs.cetis.ac.uk/adam/2013/05/03/learning-analytics
          <source>-interoperability. Accessed 2 November 2018</source>
        </mixed-citation>
      </ref>
      <ref id="ref10">
        <mixed-citation>
          10.
          <string-name>
            <surname>David</surname>
            <given-names>R.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Sandra</surname>
            <given-names>L.</given-names>
          </string-name>
          (
          <year>2005</year>
          )
          <article-title>Serious games: Games that educate, train, and inform</article-title>
          . ACM
        </mixed-citation>
      </ref>
      <ref id="ref11">
        <mixed-citation>
          11.
          <string-name>
            <surname>Ding</surname>
            <given-names>C</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Li</surname>
            <given-names>T</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Peng</surname>
            <given-names>W</given-names>
          </string-name>
          (
          <year>2008</year>
          )
          <article-title>On the equivalence between non-negative matrix factorization and probabilistic latent semantic indexing</article-title>
          .
          <source>Computational Statistics &amp; Data Analysis</source>
          <volume>52</volume>
          :
          <fpage>3913</fpage>
          -
          <lpage>3927</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref12">
        <mixed-citation>
          12.
          <string-name>
            <surname>Dorribo-Camba</surname>
            <given-names>J</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Alducin-Quintero</surname>
            <given-names>G</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Perona</surname>
            <given-names>P</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Contero</surname>
            <given-names>M</given-names>
          </string-name>
          (
          <year>2013</year>
          )
          <article-title>Enhancing model reuse through 3D annotations: a theoretical proposal for an annotation-centered design intent and design rationale communication</article-title>
          . American Society of Mechanical Engineers
        </mixed-citation>
      </ref>
      <ref id="ref13">
        <mixed-citation>
          13.
          <string-name>
            <surname>Grant</surname>
            <given-names>S</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Cordy</surname>
            <given-names>JR</given-names>
          </string-name>
          <source>(eds)</source>
          (
          <year>2010</year>
          )
          <article-title>Estimating the optimal number of latent concepts in source code analysis</article-title>
          .
          <source>IEEE</source>
        </mixed-citation>
      </ref>
      <ref id="ref14">
        <mixed-citation>
          14.
          <string-name>
            <surname>Hucka</surname>
            <given-names>M</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Graham MJ</surname>
          </string-name>
          (
          <year>2018</year>
          )
          <article-title>Software search is not a science</article-title>
          ,
          <source>even among scientists</source>
          <volume>141</volume>
          :
          <fpage>171</fpage>
          -
          <lpage>191</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref15">
        <mixed-citation>
          15. Ioan Marius Bilasco, Jérôme Gensel, Marlène Villanova-Oliver, Hervé
          <string-name>
            <surname>Martin</surname>
          </string-name>
          (
          <year>2006</year>
          )
          <article-title>An MPEG-7 framework enhancing the reuse of 3D models</article-title>
          :
          <fpage>65</fpage>
          -
          <lpage>74</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref16">
        <mixed-citation>
          16.
          <string-name>
            <surname>Jackson</surname>
            <given-names>C</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Buxton</surname>
            <given-names>M</given-names>
          </string-name>
          (
          <year>2007</year>
          )
          <article-title>The design reuse benchmark report: Seizing the opportunity to shorten product development</article-title>
        </mixed-citation>
      </ref>
      <ref id="ref17">
        <mixed-citation>
          17.
          <string-name>
            <surname>Kelly</surname>
            <given-names>H</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Howell</surname>
            <given-names>K</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Glinert</surname>
            <given-names>E</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Holding</surname>
            <given-names>L</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Swain</surname>
            <given-names>C</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Burrowbridge</surname>
            <given-names>A</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Roper</surname>
            <given-names>M</given-names>
          </string-name>
          (
          <year>2007</year>
          )
          <article-title>How to build serious games</article-title>
          .
          <source>Communications of the ACM</source>
          <volume>50</volume>
          :
          <fpage>44</fpage>
          -
          <lpage>49</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref18">
        <mixed-citation>
          18.
          <string-name>
            <surname>Ko</surname>
            <given-names>AJ</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Myers</surname>
            <given-names>BA</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Coblenz</surname>
            <given-names>MJ</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Aung</surname>
            <given-names>HH</given-names>
          </string-name>
          (
          <year>2006</year>
          )
          <article-title>An exploratory study of how developers seek, relate, and collect relevant information during software maintenance tasks</article-title>
          .
          <source>IEEE Transactions on software engineering</source>
          <volume>32</volume>
          :
          <fpage>971</fpage>
          -
          <lpage>987</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref19">
        <mixed-citation>
          19.
          <string-name>
            <surname>Loh</surname>
            <given-names>CS</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Sheng</surname>
            <given-names>Y</given-names>
          </string-name>
          , Ifenthaler D (eds) (
          <year>2015</year>
          )
          <article-title>Serious games analytics: Theoretical framework</article-title>
          .
          <source>Serious games analytics</source>
          . Springer
        </mixed-citation>
      </ref>
      <ref id="ref20">
        <mixed-citation>
          20.
          <string-name>
            <surname>Mamykina</surname>
            <given-names>L</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Manoim</surname>
            <given-names>B</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Mittal</surname>
            <given-names>M</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Hripcsak</surname>
            <given-names>G</given-names>
          </string-name>
          , Hartmann B (eds) (
          <year>2011</year>
          )
          <article-title>Design lessons from the fastest q&amp;a site in the west</article-title>
          .
          <source>Proceedings of the SIGCHI conference on Human factors in computing systems. ACM</source>
        </mixed-citation>
      </ref>
      <ref id="ref21">
        <mixed-citation>
          21.
          <string-name>
            <surname>McShaffrey</surname>
            <given-names>M</given-names>
          </string-name>
          (
          <year>2005</year>
          )
          <article-title>Game coding complete</article-title>
          .
          <source>Scene-graphs. Paraglyph Inc</source>
        </mixed-citation>
      </ref>
      <ref id="ref22">
        <mixed-citation>
          22.
          <string-name>
            <surname>Mestadi</surname>
            <given-names>W</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Nafil</surname>
            <given-names>K</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Touahni</surname>
            <given-names>R</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Messoussi</surname>
            <given-names>R</given-names>
          </string-name>
          (
          <year>2018</year>
          )
          <article-title>An Assessment of Serious Games Technology: Toward an Architecture for Serious Games Design</article-title>
          .
          <source>International Journal of Computer Games Technology 2018</source>
        </mixed-citation>
      </ref>
      <ref id="ref23">
        <mixed-citation>
          23.
          <string-name>
            <surname>Millington</surname>
            <given-names>I</given-names>
          </string-name>
          (
          <year>2007</year>
          )
          <article-title>Game Physics Engine Development</article-title>
          . CRC Press
        </mixed-citation>
      </ref>
      <ref id="ref24">
        <mixed-citation>
          24.
          <string-name>
            <surname>Millington</surname>
            <given-names>I</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Funge</surname>
            <given-names>J</given-names>
          </string-name>
          (
          <year>2009</year>
          )
          <article-title>Artificial intelligence for games</article-title>
          . CRC Press
        </mixed-citation>
      </ref>
      <ref id="ref25">
        <mixed-citation>
          25.
          <string-name>
            <surname>Murphy-Hill</surname>
            <given-names>E</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Zimmermann</surname>
            <given-names>T</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Nagappan</surname>
            <given-names>N</given-names>
          </string-name>
          (
          <year>2014</year>
          )
          <article-title>Cowboys, ankle sprains, and keepers of quality: How is video game development different from software development?:1-11</article-title>
        </mixed-citation>
      </ref>
      <ref id="ref26">
        <mixed-citation>
          26.
          <string-name>
            <surname>Petridis</surname>
            <given-names>P</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Dunwell</surname>
            <given-names>I</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Panzoli</surname>
            <given-names>D</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Arnab</surname>
            <given-names>S</given-names>
          </string-name>
          , Protopsaltis, Aristidis and Hendrix, Maurice, Freitas S de (
          <year>2012</year>
          )
          <article-title>Game engines selection framework for high-fidelity serious applications</article-title>
          .
          <source>International Journal of Interactive Worlds</source>
        </mixed-citation>
      </ref>
      <ref id="ref27">
        <mixed-citation>
          27.
          <string-name>
            <surname>Rosen</surname>
            <given-names>C</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Shihab</surname>
            <given-names>E</given-names>
          </string-name>
          (
          <year>2016</year>
          )
          <article-title>What are mobile developers asking about? a large scale study using stack overflow</article-title>
          .
          <source>Empirical Software Engineering:</source>
          <fpage>1192</fpage>
          -
          <lpage>1223</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref28">
        <mixed-citation>
          28.
          <string-name>
            <surname>Sadowski</surname>
            <given-names>C</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Stolee</surname>
            <given-names>KT</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Elbaum</surname>
            <given-names>S</given-names>
          </string-name>
          (
          <year>2015</year>
          )
          <article-title>How developers search for code: a case study</article-title>
          .
          <source>Proceedings of the 2015 10th Joint Meeting on Foundations of Software Engineering:</source>
          <fpage>191</fpage>
          -
          <lpage>201</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref29">
        <mixed-citation>
          29.
          <string-name>
            <surname>Singer</surname>
            <given-names>J</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Lethbridge</surname>
            <given-names>T</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Vinson</surname>
            <given-names>N</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Anquetil</surname>
            <given-names>N</given-names>
          </string-name>
          (
          <year>2010</year>
          )
          <article-title>An Examination of Software Engineering Work Practices: CASCON First Decade High Impact Papers</article-title>
        </mixed-citation>
      </ref>
      <ref id="ref30">
        <mixed-citation>
          30.
          <string-name>
            <surname>Söbke</surname>
            <given-names>H</given-names>
          </string-name>
          , Seicher A (eds) (
          <year>2016</year>
          )
          <article-title>Serious Games Architectures and Engines</article-title>
          .
          <source>Entertainment Computing and Serious Games</source>
          . Springer
        </mixed-citation>
      </ref>
      <ref id="ref31">
        <mixed-citation>
          31.
          <string-name>
            <surname>van der Vegt</surname>
            <given-names>W</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Nyamsuren</surname>
            <given-names>E</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Westera</surname>
            <given-names>W</given-names>
          </string-name>
          (
          <year>2016</year>
          )
          <article-title>RAGE reusable game software components and their integration into serious game engines</article-title>
          :
          <fpage>165</fpage>
          -
          <lpage>180</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref32">
        <mixed-citation>
          32.
          <string-name>
            <surname>Vasudevamurt</surname>
            <given-names>VB</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Uskov</surname>
            <given-names>A</given-names>
          </string-name>
          (
          <year>2015</year>
          )
          <article-title>Serious Game Engines: Analysis and Application</article-title>
          .
          <source>Electro/Information Technology (EIT)</source>
          ,
          <year>2015</year>
          IEEE International Conference on
        </mixed-citation>
      </ref>
      <ref id="ref33">
        <mixed-citation>
          33.
          <string-name>
            <surname>Wallach</surname>
            <given-names>HM</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Murray</surname>
            <given-names>I</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Salakhutdinov</surname>
            <given-names>R</given-names>
          </string-name>
          , Mimno D (eds) (
          <year>2009</year>
          )
          <article-title>Evaluation methods for topic models</article-title>
          . ACM
        </mixed-citation>
      </ref>
      <ref id="ref34">
        <mixed-citation>
          34.
          <string-name>
            <surname>Westera</surname>
            <given-names>W</given-names>
          </string-name>
          ,
          <string-name>
            <surname>van der Vegt</surname>
            <given-names>W</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Bahreini</surname>
            <given-names>K</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Dascalu</surname>
            <given-names>M</given-names>
          </string-name>
          ,
          <string-name>
            <surname>van Lankveld</surname>
            <given-names>G</given-names>
          </string-name>
          (
          <year>2016</year>
          )
          <article-title>Software Components for Serious Game Development</article-title>
        </mixed-citation>
      </ref>
      <ref id="ref35">
        <mixed-citation>
          35.
          <string-name>
            <surname>Wim</surname>
            <given-names>V</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Westera</surname>
            <given-names>W</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Nyamsuren</surname>
            <given-names>E</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Georgievand</surname>
            <given-names>A</given-names>
          </string-name>
          (
          <year>2016</year>
          )
          <article-title>RAGE architecture for reusable serious gaming technology components 2016:3</article-title>
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>