=Paper=
{{Paper
|id=Vol-3147/paper2
|storemode=property
|title=Advantages and challenges of extracting process knowledge through serious games
|pdfUrl=https://ceur-ws.org/Vol-3147/paper2.pdf
|volume=Vol-3147
|authors=Thomas Schemmer,Jenny Reinhard,Philipp Brauner,Martina Ziefle
|dblpUrl=https://dblp.org/rec/conf/gamifin/SchemmerRBZ22
}}
==Advantages and challenges of extracting process knowledge through serious games==
https://ceur-ws.org/Vol-3147/paper2.pdf
Advantages and challenges of extracting process knowledge
through serious games
Thomas Schemmer1, Jenny Reinhard1, Philipp Brauner1 and Martina Ziefle1
1 Human-Computer Interaction Center, Campus Boulevard 57, RWTH Aachen University, 52074 Aachen,
Germany
Abstract
Digitalization promises huge improvements in various domains, such as production, health
care, or mobility, through the integration of big data and artificial intelligence (AI). However,
AI often builds on labelled data but labeling data can be complex or expensive, depending on
both the properties of the data and access to people with domain knowledge. In particular, an
underexplored field is capturing process knowledge, i.e., knowledge about the relationships
among process steps. In this work, we propose and evaluate a game-based approach for
capturing process knowledge. Taking the cooking domain as an example, we developed a
prototype, in which players act as chef and cook dishes following their own recipes while each
action is logged. The captured data is then compared to ground-truth models of common
recipes. While the quantitative evaluation shows a decrease in motivation as well as fewer
logged steps, qualitative feedback from participants identifies possible improvements of the
concept. In summary, games can be a suitable approach for extracting experts’ process
knowledge, when certain user requirements are considered.
Keywords1
Process knowledge, knowledge harvesting, process mining, knowledge extraction, domain
expertise, serious games
1. Introduction relationships between different entities are of
interest. A question in this area is if the
digitization of knowledge can be improved by
Sustainable knowledge management is a key
amplifier concepts such as gamification or serious
topic in numerous domains, such as
games in terms of the amount of data or data
production [1], [2], health care [3], and
quality and whether this can be linked to
management [4]. One particular question is how
individual user characteristics. Although our
(expert) knowledge can be systematically
research addresses the extraction of expert
captured digitally so that it can later be used as a
knowledge from process planning in
knowledge base, for training, or for the creation of
manufacturing in the long term, here we consider
data-driven decision support systems [5]–[7].
process knowledge that many people have: The
While capturing specific types of knowledge is
preparation of food with the recipes as
easy and can build on a vast pool of novices (for
manifestations of their process knowledge. At a
example, massive image classification via
later stage, we will transfer our concept and
MTurk [8]), capturing expert knowledge becomes
findings to the production domain.
hard when access to experts is limited, expensive,
or the tasks to be captured are complex [9].
In this paper, we consider the special case of 1.1. Background
capturing domain-specific process knowledge,
i.e., when not individual items need to be In the long run, we aim at generating a digital
classified or labelled but when also the representation of the process knowledge from
6th International GamiFIN Conference 2022 (GamiFIN 2022),
April 26-29, 2022, Finland
EMAIL: lastname@comm.rwth-aachen.de (A. 1 – A.4);
ORCID: 0000-0002-8584-0126 (A. 1); 0000-0003-2837-5181 (A.
3); 0000-0002-6105-4729 (A. 4)
©️ 2022 Copyright for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
11
experienced process planners in textile can then be used to train AI models for automation
engineering. On the one hand, this sector is or decision support.
characterized by domain experts that have much In our serious game, the worker is intended to
tacit experiential knowledge or even knowledge play inside a gamified version of the shop floor,
in motor memory. On the other hand, most where all tools, machines and resources are
companies in the sector are often reluctant to available. As before, the player then gets
exploit the opportunities offered by digitization prompted to manufacture certain products, while
and digital knowledge management [10]. the game tracks his actions.
Consequently, the potential of capturing and then However, as this field is not yet researched, we
using digital knowledge for training or building conducted a proof-of-concept study, providing
automated decision support systems is untapped. first insights on the pros and cons of our approach.
Currently, process planning is more manual To be able to reach more participants for the first
than digital: Planners usually write down their proof-of-concept study, we realized a game for
executed steps for a certain production process on extracting cooking process knowledge instead of
paper, shortly after manufacturing the the specialized industrial use case. This allows us
product [10]. Normally, this only includes the to gather extensive feedback more quickly,
steps taken and not the reasoning behind the without the need for experts with their specific
decisions. To make these textual artifacts usable domain knowledge. The core idea should then be
for building training materials, knowledge bases, transferable to production use-cases, such as
or for training an AI, they must be digitalized and textile engineering, in the future.
formalized. Yet, this is cumbersome and error- Compared to previous approaches, this would
prone for the workers, as many modern tools that have multiple advantages. One, the knowledge is
are used within production settings, such as Excel, immediately available, so the digitalization and
are confusing due to poor user experience and unification would be simpler, faster, and more
complexity [11]. Also, multiple workers will note accurate. Two, serious games have shown an
down information differently, so the resulting increase in motivation, which would favour the
digitalized information must be unified. workers. The increased motivation could lead to
Another approach for gathering the required increased productivity, benefiting the companies.
process knowledge might be to interview workers Three, the time spent gathering the logs could be
to formulize plans for several different products reduced, as all input will be stored in one place.
and ask for their reasoning in interviews [9]. Yet,
this would be more cumbersome, as this would 2. Related work
require additional staff for conducting the
interviews, the interviews would have a limited
This chapter introduces the core concepts of
time frame and would thus require focusing on the
our vision and relates these to existing research.
most important or difficult cases only [9].
Both approaches face two difficulties. First,
they require the worker to work in a repetitive 2.1. Serious games
setting, which reduces internal motivation [12].
Second, the data would need to be digitalized, A Serious Games (SG) is a (often computer-
which would require human classification and mediated) game whose goal is not primarily
domain expert knowledge, and additional entertainment, but that convey knowledge or
computational overhead. behaviour change [13]. They usually use
simplified abstractions of problems and are thus
1.2. Vision and approach not necessarily complete [14]. In our case, we
would build on the persuasive potential of
games [15] to motivate people to share domain
As a solution for these problems, we propose
specific process knowledge. Note that SG differ
serious games as a method for extracting
from gamification, where unaltered activities are
industrial process knowledge. Experts would
reinforced with game elements, such as timers,
playfully interact with a (simulated) production
points, badges, or leaderboards [16]–[18].
environment and thus share their experience and
expertise with a system that captures all Both gamification and SG have shown success
in medical contexts, (e.g., reminding people to
interactions. The knowledge captured digitally
wash hands properly [19], [20]), personal
education (e.g., increased learning of a new
12
language [21] or to nudge students to learn mixed results, as tasks were completed faster but
efficiently [22].), but also in production (e.g., to also failure rates increased.
convey knowledge and to study human behaviour As a reverse, [31] and [32] used gamification
in supply chains [23]). elements to facilitate learning in an industrial
setting, either to teach lean manufacturing, or to
2.2. Knowledge extraction and identify warning indicators. While the authors
used gamification in an industrial use-case, they
process mining focused on learning for the user, not extracting
knowledge from them.
Knowledge Extraction (KE) is the act of This overview highlights the missing research
gathering knowledge about a topic from into combining PM and gamification. Both have
structured sources, such as databases or XML, or previously shown benefits on their own, but only
unstructured sources, such as texts, images or—as rarely together. Especially in the industrial use-
in our case—games. The main goal is to create a case, where PM is widely used, the lack of a
ruleset or history for an AI to reason upon, to combined approach is glaring.
accurately predict solutions for the future. A very
common approach is to create triplets, which are
small information bits, linking multiple topics to
2.3. Motivation
each other. If enough triplets are created, one can
follow this reasoning chain to create new The major benefit for the workers would be
information. This concept was the basis for the higher hedonic motivation while sharing
creation of the reasoner pellet [24]. KE is also knowledge. Psychology divides motivation into
used in medicine, either to provide data for dietary intrinsic ("I work on this topic because it is fun.")
recommender systems [25] or to scrape patient and extrinsic ("I work on this topic because I get
information from clinical data [26]. paid for it") motivations [33]. Here, intrinsic
While the above examples all focus on creating motivation is more important, as extrinsic
rulesets, Process Mining (PM) is working towards motivation quickly degrades and tasks are not
a unified process model [27]. This model can then continued if the rewards decrease.
be used to compare it with running work iterations How can motivation be measured? Motivation
or reasoned from. PM extracts information from can either be measured by using psychometric
event logs, which is a collection of activities, scales or by observing behaviour. Regarding the
together with timestamps and process identifiers. former, the Situational Motivation Scale (SIMS)
PM defines a process as a theoretical series of is a validated scale that measures four dimensions
activities (or actions of the worker), whereas a of motivation, namely Intrinsic Motivation,
specific execution of this process is called a trace. Identified Regulation, External Regulation and
Similar traces are grouped together, creating a Amotivation [34]. For the latter, the Free-Choice
variant, which in turn are used to create the model. Measurement (FCM) can be used: Without any
PM also defines several disparity measurements external control people can do a task or interact
between a variant and the model [28]. PM is with a system. The time people invest is then an
widely used in business, as their production log is indicator of a persons’ motivation [35].
the ideal candidate to reason upon [27]. Combining both, SIMS and FCM, will provide
Computing a ruleset from a given dataset is richer reasoning behind the users’ behaviour.
difficult, as a wide variety of individual deviations
as well as unification must be considered. 2.4. Research gap and objective
The combination of PM and gamification is
promising, as they complement each other. This The extraction of process knowledge has been
has been done in some cases, but not many in an insufficiently solved so far. Serious games
industrial setting. For example, [29] used PM to promise to motivate people to interact longer in a
classify data collected from a gamified virtual environment and thus make capturing their
experiment. We on the other hand would like to process knowledge possible by logging their
use gamification to create a better process log. In interactions. In this paper we investigate if process
contrast, [30] created a gamified environment in a knowledge can be captured by means of a SG,
production setting, but without extracting or whether a SG achieves better results than a control
analyzing knowledge. Their evaluation showed condition, and what role user diversity and
13
motivation play. Our research is guided by the their interactions in the virtual kitchen (i.e., to
following hypotheses: extract process knowledge in the cooking
H0: Process knowledge can be captured by domain). To achieve this, a prompt displays only
means of a SG. the name of the dish and the player is then given
As SG are often suggested as being more interaction opportunities to perform the steps
motivating, we compare the SG with a he/she would take to cook the dish in real life.
functionally equivalent control condition and Each interactable component in the game is
postulate: modelled as either distinct cupboards, crates, or
H1: Users of the serious game for knowledge machinery. Cupboards hold container, i.e., pots
harvesting report higher motivation than users of and pans. Crates contain ingredients and
a control environment. machinery is e.g., an oven. Each container can
H2: User factors influence reported motivation hold an infinite amount of ingredients to reflect
after of the serious game the different steps of a recipe, such as adding
SG promise higher motivation and higher tomatoes. The container, and therefore the
motivation goes hand in hand with higher contained ingredients, can be cooked, baked and
performance. Therefore, the following two seasoned. The ingredients are divided into dairy
hypothesis address the products (milk, cheese, eggs), meats (fish, beef,
H3: A serious game captures more process minced beef), carbohydrates (noodles, bread) and
knowledge compared to the control condition. vegetables (paprika, onions). Each category is
H4: A serious game provides more accurate contained in its own crate or inside a fridge. We
process knowledge compared to the control choose these ingredients to allow for many
condition. possible recipes. Additionally, some of these
H5: Higher motivation leads to more accurate ingredients can be cut into smaller pieces.
process knowledge that can be captured. There are three distinct forms of interaction of
Hypotheses H1 and H2 focus on the users’ increasing complexity: cutting ingredients,
motivation, whereas H3 and H4 address the seasoning and cooking. Cutting ingredients will
benefit of KE by means of a SG. H5 connects both always result in the same outcome without any
aspects, providing pointers for further research. choice of the player. Seasoning recipes have a
wider variety of choices, but it is generally
3. Implementation of conditions understood to have only a small effect on the
result. This is different to cooking, as—depending
on the heat and time settings—it is possible to
To evaluate the feasibility of process KE by burn dishes in real life. To keep the complexity of
means of SG, we implemented a low-poly kitchen the game low, burning dishes is not possible in the
game using the Unity3D engine. We used WebGL game. Figure 2 displays the user interface for
to make the game accessible to participants using interacting with the stove. The player can choose
a browser, featuring keyboard and mouse input. It the heat level, as well as the duration and can see
is designed as a top-down, fixed-perspective a preview of the current ingredients.
camera. Participants play a chef interacting with
the different components of the kitchen. Figure 1 Stove
shows a screenshot of the game with the chef
walking to the fridge.
Figure 2: User interface for cooking dishes.
Figure 1: Screenshot of the Game.
There are two kinds of recipe queries in the
game: mandatory and free choice. Free choice
The goal of the game is to extract the recipes recipes are not logged, and players can decide
for several dishes from the players by capturing how many recipes they want to complete. Only
14
the amount of completed free choice recipes will
be used as a metric. Conversely, to complete an
experiment each player must recreate the five
mandatory dishes as recipes in the game. The
recipes are green salad, omelettes, greek salad,
burger and spaghetti bolognese. All interaction
for these recipes is logged into a database,
creating a process log. This allows a direct
analysis of the resulting process models with the
help of PM tools.
PM allows for either the recreation of a process Figure 4: Interface of the control condition. Using
model from a sufficient log, or conformance a drag and drop interface the participants could
checking the log with a ground truth model. We prepare selected dishes.
have chosen the latter, as creating an accurate
model would require hundreds of traces, which
will not be feasible for early evolution of the
concept. We have therefore created ground truth 4. Evaluation
models for each of the mandatory recipes. The
ground truth model for a burger is depicted in To evaluate the general feasibility of our
Figure 3. This model follows standard PM approach, we conducted a user study with the SG
notation. Rounded rectangles represent different and a control group. The following sections
activities, + denotes an AND transition, whereas x present our experimental method, the sample, and
denotes an XOR transition. Note that this model the main results of the study.
allows multiple vegetables by heaving a loop.
4.1. Method
MincedMeat Cook
+ CUTSalad + HandIn The participants of our study were randomly
assigned to either the SG or the control condition
x CUTTomato x
(game type as a between-subject factor). The
CUTBread control group is introduced to a bland drag and
Figure 3: PM model for the recipe burger. drop interface. Both groups have the same
interaction possibilities and target recipes and
were exclusively played on a computer.
3.1. Control condition Participants were recruited from friends and the
websites Positly and PollPool during May 2021.
We created an additional, functionally Due to the pandemic restrictions, they were able
equivalent, drag-and-drop web interface as a to choose their own place to partake in the
control setting. This interface was intentionally experiment.
designed in a bland, unenticing way to reflect the As independent variables, we collected the
visuals of modern tools such as Excel. It does not participants’ demographics using an online survey
include any form of gamification. All interactions on Qualtrics, such as age, sex, as well as job type
and resources that are available in the cooking SG and -field. To further evaluate the influence of the
are also available in the control condition. Error! effect of exploratory user factors, we further
Reference source not found. depicts the measured the participants’ attitudes towards
interface. technology [36] and their attitude towards games
on 5-point Likert scales under the assumption that
experienced players might evaluate the game
differently than people who don’t enjoy playing
games. Cronbach's α shows that both scales have
a high internal consistency (gaming α=.916,
attitude towards technology α=.911).
As dependent variables we measured a) the
participants’ motivation after the interaction using
the SIMS scale (intrinsic motivation α=.946,
15
internal regulation α=.862, external regulation Can process knowledge be captured by
α=.837, amotivation α=.811), b) the number of means of a serious game? First, we investigated
process steps done for a recipe, and c) the quality whether process knowledge can be generated
of the recipes cooked by the participants measured from the interaction logs of both the SG and the
by PM’s fitness measure. For the last two control condition and what the quality of the
measures, log files captured the interactions with captured process knowledge is.
the system and thus the process steps while Across the five different recipes from the
preparing the dishes. experiment, the participants performed on
average 11 steps per recipe (see Figure 8). The
4.2. Description of the sample resulting average fitness is .51 and thus
satisfactory, with the fitness of the captured
process model for the green salad being highest
Overall, 60 people participated in the study, 21 and for spaghetti being lowest).
in the SG (34%) and 39 in the control condition
(64%). Most of our participants were in the age Is the serious game more motivating than
range between 18–30 years and most of the
the control condition? To compare the reported
participants were women (61%). In terms of their Intrinsic Motivation between both conditions, we
current employment, our sample was diverse, calculated a Mann-Whitney U (MW-U) test.
with participants working in technical and non- Although the median Intrinsic Motivation appears
technical domains (see Figure 6). lower for the SG condition (md=3.5) than for the
control condition (md=4.7), this difference is not
statistically significant (p=.223>.05). Therefore,
H1 is not supported by the evidence.
Note the non-normal distribution, measured by
a seven-point Likert scale, of intrinsic motivation
for the SG, displayed in Figure 7. While the
control group has a central peak at around 4.9, the
SG version has two peaks (bimodal distribution)
at 2 ("Didn't enjoy it") and 6 ("Did enjoy it").
Figure 5: Age and gender distribution of the
participants (male=blue, female=red).
Figure 7: Histogram for Intrinsic Motivation after
the serious game (red) and control condition
(blue), measured on a 7-point Likert scale.
Figure 6: Job distribution of the participants.
Do individual user-factors influence
5. Results motivation after interacting with the serious
game? We surveyed six different user factors in
In the following, the results of the experiment this study: Gaming disposition, attitude towards
are presented in the order of the hypotheses. technology, job field, highest degree, gender, and
age. Neither job field, nor highest academic
degree, nor the participants’ gender had any
16
significant influence on intrinsic motivation
(p’s≫.05).Error! Reference source not found.
As both gaming and attitude towards technology
are continuous measurements, we evaluated their
relation to the intrinsic motivation with a linear
regression. In neither of the games does Gaming
disposition have a significant influence (serious
game: p=.411, control: p=.086). On the other
hand, attitude towards technology, has a
significant influence on the SG version
(p=.042<.05, est. β=-2.09, SE=.959, t=-2.18,
R2=0.2). As this effect is negative, we conclude
that participants with higher attitude towards
technology found the SG less motivating. Figure 9: Free Choice Measurement for both SG
(red) and Control (blue)
Does the serious game capture more process
data? Two measurements were analyzed to Does the serious game provides more accurate
evaluate the validity of the H3. First, we compared process knowledge? As H3 evaluated the amount
the number of steps per recipe (see Figure 8). In of data and not the quality thereof, we measured
the control condition, the participants contributed the difference in quality according to the models
on average 16 recipe steps compared to 11 in the we provided.
SG condition. A MW-U test showed that this To evaluate the accurateness of the
difference is significant (p<.001). Consequently, participants’ recipes, the standard measurement in
H3 is refuted. PM fitness was used. Figure 10 depicts the
The second measurement is the Free-Choice average fitness for each of the recipes in both
Measurement. Figure 9 depicts the histogram for versions. Here, the overall difference as measured
both versions. As both versions are non-normally by Welchs’ t-test is not significant (t(4)=-0.878,
distributed, we calculated a MW-U-test and there p=.406). Thus, the accurateness of the data
is no significant difference between both versions acquired in the serous games is not higher
(p=.216) (n(C)=39, n(SG)=21, mdn(C)=0, compared to the control condition and H4 is not
mdn(SG)=0),. Therefore, the serious game does supported.
not provide more data compared to the control
condition and H3 is discarded.
Figure 10: Fitness per Recipe for both SG (red)
and Control (blue), allowing all Ingredients.
Figure 8: Steps per Recipe for both SG (red) and Does higher motivation of the participants
Control (blue) yield more accurate process knowledge
captured? Next, we analyse if the partcipants’
motivation relates to the accuracy of the captured
process knowledge. We first consider the SG
condition and then the control condition.
17
We compared the averaged fitness of all quite decent fitness. Thus, our SG approach for
recipes from each participant with the extracting process knowledge worked well.
participants’ SIMS. In the SG condition, no However, in the end, none of our formulated
correlations between the averaged Fitness and research hypotheses that compared the SG against
Intrinsic Motivation (p=.252, R2=.068), a conventional user interface could be validated.
Identified Regulation (p=.239, R2=.072), There was no significant difference in motivation
External Regulation (p=.720, R2=.007) or (H1, as measured by the SIMS) between the
Amotivation (p=.204, R2=.083) from the SIMS playful SG and the rather dull control condition.
scales were found. Thus, motivation was not As we targeted the SG towards the elderly
linked to the accuracy of the captured process workers, different user factors have been
knowledge in the SG condition. discussed (H2). Yet, there hasn't been a significant
Contrary, there was a significant negative influence from age, gender, job field, or gaming
influence of both External Regulation disposition. Only attitude towards technology had
(p=.009<.05, est. β=-2.06, R2=.185) and a negative effect.
Amotivation (p=.009<.05, est. β=-2.48, R2=.185) While there was a significant difference in the
on the average Fitness for the control condition , players’ intrinsic motivation, we could not yet
but no influence of Intrinsic Motivation (p=.234, identify the specific reasons for this effect. We
R2=.058) and Identified Regulation (p=.324, found however that—independent of the
R2=.022). experimental condition—older players reported a
Thus, the findings suggest that motivation higher intrinsic motivation. This finding suggests
influences the accuracy of the captured process that older participants might be more willing to
knowledge only in the control condition but not in share their experiences. A potential for KE and
the SG condition. Consequently, H5 is partially management that should be taped.
supported although a more thorough investigation Additionally, the SG provided less additional
with a larger sample size is needed. data (H3) and no difference in data accuracy (H4).
The only measurable difference between the two
6. Discussion versions was the better usage of the cutting board,
as nearly every player in the SG used it, while not
even half of the control groups’ players used it.
In this article, we presented the rationale for We attribute this to a more intuitive understanding
capturing process knowledge and a SG situated in and higher visual clarity of the cutting board
a kitchen environment that aims at extracting compared to the text field in the control condition.
recipes as one of the most common manifestations Due to its significant effort to program the SG
of process knowledge that most people have. The compared to the conventional interface, we
overall goal was to let the players create their own currently cannot recommend the SG in its current
recipes for a set of dishes and compare the results form to rise the workers’ motivation, or to
with a ground truth. We wanted to analyze if a increase the amount or quality of the extracted
SGs approach provides two major benefits: process knowledge.
Firstly, it should increase the motivation of the As we were only able to show a negative effect
player, because it would be more interactive than of External Regulation and Amotivation on the
the blander counterpart. This would have been a control version (H5), we suggest evaluating this
major benefit to the workers. Secondly, deriving
difference further. The lack of negative influence
from this increase in motivation, players should of these modifiers in the SGs is an interesting
have created additional data, as well as have a point for further investigations.
higher accuracy of their recipes. This has been
evaluated in an online experiment with a control
group that used a functionally equivalent drag and 7. Limitations, outlook, implications
drop interface for sharing recipes. Next, we
discuss the findings of our experiment and Of course, this study is not without limitations.
provide pointers for further research. The biggest limitation is certainly the small
First, our results indicate that we can extract sample size, which limits the transferability and
peoples’ cooking knowledge for five common the consideration of user diversity effects. Also,
recipes in our study. The generated process logs we found that the steeper learning curve of the
were analyzed with PM metrics and achieved serious game led to more dropouts compared to
the control condition and thus unequal group
18
sizes. Nevertheless, the findings suggests that valuable feedback. We thank the anonymous
process knowledge can be captured digitally reviewers for their immensely valuable feedback.
through serious games and that individual
motivation, as a facet of user diversity, influences 9. References
the result quantity and quality. This needs to be
investigated and modelled in more detail in future
[1] M. Fakhar Manesh, M. M. Pellegrini, G.
studies. Ideally also under laboratory conditions
Marzi, and M. Dabic, “Knowledge
and as a within-subject experiment, to mitigate the
Management in the Fourth Industrial
various biases of online survey.
Revolution: Mapping the Literature and
A major downside of the online evaluation
Scoping Future Avenues,” IEEE
approach was the difficulty of learning the basic
Transactions on Engineering Management,
interaction with the game. Additional feedback
vol. 68, no. 1, pp. 289–300, Feb. 2021.
provided by the participants centered around
[2] P. Brauner et al., “A Computer Science
confusion about the game and its interactivity.
Perspective on Digital Transformation in
While we provided a text-based tutorial, many
Production,” ACM Transactions on Internet
participants didn't truly understand the SG, which
lead to indecision and quitting the game. This of Things, vol. 3, no. 2, pp. 1–32, May 2022.
[3] C. El Morr and J. Subercaze, “Knowledge
difficulty resulted in a small sample size for the
Management in Healthcare,” in Handbook of
SG, which limits the overall validity of our
Research on Developments in E-Health and
findings. For further studies we thus need to
Telemedicine, IGI Global, pp. 490–510.
flatten the learning curve, for example through
[4] P. Drucker, The Age of Discontinuity -
appropriate tutorials, We also recommend a
Guidelines on Our Changing Society. New
supplementary, qualitative experiment in which
York, NY, USA: Heinemann, 1968.
the participants get a live explanation and training
[5] F. Kimura, H. Ariyoshi, H. Ishikawa, Y.
session before being asked to provide the recipes.
Naruko, and H. Yamato, “Capturing Expert
This could significantly reduce the participants’
Knowledge for Supporting Design and
confusion and might result in a significant change
Manufacturing of Injection Molds,” CIRP
in the overall results. Of course, this would mean
Annals, vol. 53, no. 1, pp. 147–150, 2004.
an even further increase in workload, compared to
[6] P. Fantini, M. Pinzone, F. Sella, and M.
a drag-and-drop or Excel-based solution.
Taisch, “Collaborative Robots and New
In summary, this study showed that process
Product Introduction: Capturing and
knowledge from the commonly known domain of
Transferring Human Expert Knowledge to
cooking can be captured with a serious game,
the Operators,” in International Conference
even if the consideration of motivational aspects
on Applied Human Factors and Ergonomics,
revealed few surprises. While the digital capturing
2017, pp. 259–268.
of cooking knowledge itself is only of marginal
[7] M. Mansfield, V. Tamma, P. Goddard, and
interest in specialized areas (e.g., to make cultural
F. Coenen, “Capturing Expert Knowledge
differences in the preparation of food measurable
for Building Enterprise SME Knowledge
or for saving cultural heritage), the findings
Graphs,” in Proceedings of the 11th on
suggest the transferability of this concept to other
domains and contexts. We postulate that this Knowledge Capture Conference, 2021, pp.
129–136.
approach enables capturing process knowledge in
[8] P. Welinder and P. Perona, “Online
areas of manufacturing that have been little
crowdsourcing: Rating annotators and
digitised so far that may serve as data to increase
obtaining cost-effective labels,” in IEEE
automation and provide decision support [2], [10].
Computer Society Conference on Computer
Vision and Pattern Recognition -
8. Acknowledgements Workshops, 2010, pp. 25–32.
[9] R. R. Hoffmann, “The Problem of Extracting
Funded by the Deutsche Forschungs- the Knowledge of Experts from the
gemeinschaft (DFG, German Research Perspective of Experimental Psychology,”
Foundation) under Germany’s Excellence AI Magazine, vol. 8, no. 2, p. 53, 1987.
Strategy—EXC-2023 Internet of Production— [10] F. Brillowski, H. Dammers, H. Koch, K.
390621612. Anne Kathrin Schaar for immensely Müller, L. Reinsch, and C. Greb, “Know-
How Transfer and Production Support
19
Systems to Cultivate the Internet of Behavior, vol. 29, no. 2, pp. 345–353, Mar.
Production Within the Textile Industry,” in 2013.
International Conference on Intelligent [23] P. Brauner, R. Philipsen, and M. Ziefle,
Human Systems Integration (IHSI 2021), “Projecting Efficacy and Use of Business
2021, pp. 309–315. Simulation Games in the Production Domain
[11] P. T. Kortum and A. Bangor, “Usability Using Technology Acceptance Models,” in
Ratings for Everyday Products Measured Advances in Ergonomics of Manufacturing:
With the System Usability Scale,” Managing the Enterprise of the Future, C.
International Journal of Human-Computer Schlick and S. Trzcieliński, Eds. Springer
Interaction, vol. 29, no. 2, pp. 67–76, Jan. International Publishing, 2016, pp. 607–620.
2013. [24] E. Sirin, B. Parsia, B. C. Grau, A. Kalyanpur,
[12] C. D. Fisherl, “Boredom at Work: A and Y. Katz, “Pellet: A practical OWL-DL
Neglected Concept,” Human Relations, vol. reasoner,” Journal of Web Semantics, vol. 5,
46, no. 3, pp. 395–417, Mar. 1993. no. 2, pp. 51–53, Jun. 2007.
[13] C. C. Abt, Serious Games. Madison Books, [25] T. Eftimov, B. Koroušić Seljak, and P.
1987. Korošec, “A rule-based named-entity
[14] D. Michael and S. Chen, Serious Games: recognition method for knowledge
Games That Educate, Train and Inform. extraction of evidence-based dietary
Thomson Course Technology, 2006. recommendations,” PLOS ONE, vol. 12, no.
[15] I. Bogost, Persuasive Games: The 6, p. e0179488, Jun. 2017.
Expressive Power of Videogames. [26] G. K. Savova et al., “Mayo clinical Text
Cambridge, Massachusetts, USA: The MIT Analysis and Knowledge Extraction System
Press, 2007. (cTAKES): architecture, component
[16] D. Djaouti, J. Alvarez, J.-P. Jessel, and O. evaluation and applications,” Journal of the
Rampnoux, “Origins of Serious Games,” in American Medical Informatics Association,
Serious Games and Edutainment vol. 17, no. 5, pp. 507–513, Sep. 2010.
Applications, London: Springer London, [27] W. van der Aalst et al., “Business Process
2011, pp. 25–43. Mining: An Industrial Application,”
[17] J. Hamari, “Transforming homo economicus Information Systems, vol. 32, no. 5, pp. 713–
into homo ludens: A field experiment on 732, Jul. 2007.
gamification in a utilitarian peer-to-peer [28] W. van der Aalst et al., “Process Mining
trading service,” Electronic Commerce Manifesto,” in Business Process
Research and Applications, vol. 12, no. 4, Management Workshops, Springer, Berlin,
pp. 236–245, 2013. Heidelberg, 2012, pp. 169–194.
[18] J. Hamari, “Gamification,” in The Blackwell [29] J. A. C. Hernández, M. P. Duarte, and J. M.
Encyclopedia of Sociology, Oxford, UK: Dodero, “An architecture for skill
John Wiley & Sons, Ltd, 2019, pp. 1–3. assessment in serious games based on Event
[19] A. Higgins and M. M. Hannan, “Improved Sequence Analysis,” in Proceedings of the
hand hygiene technique and compliance in 5th International Conference on
healthcare workers using gaming Technological Ecosystems for Enhancing
technology,” Journal of Hospital Infection, Multiculturality, 2017, pp. 1–9.
vol. 84, no. 1, pp. 32–37, May 2013. [30] O. Korn, M. Funk, and A. Schmidt,
[20] P. Pereira, E. Duarte, F. Rebelo, and P. “Towards a gamification of industrial
Noriega, “A Review of Gamification for production,” in Proceedings of the 7th ACM
Health-Related Contexts,” in International SIGCHI Symposium on Engineering
Conference of Design, User Experience, and Interactive Computing Systems, 2015, pp.
Usability, 2014, pp. 742–753. 84–93.
[21] W. L. Johnson, “Serious Use of a Serious [31] M. Messaadia, A. Bufardi, J. Le Duigou, H.
Game for Language Learning,” International Szigeti, B. Eynard, and D. Kiritsis,
Journal of Artificial Intelligence in “Applying Serious Games in Lean
Education, vol. 20, no. 10, pp. 175–195, Manufacturing Training,” 2013, pp. 558–
2010. 565.
[22] J. Simões, R. D. Redondo, and A. F. Vilas, [32] R. Patriarca, A. Falegnami, A. De Nicola, M.
“A social gamification framework for a K-6 L. Villani, and N. Paltrinieri, “Serious games
learning platform,” Computers in Human for industrial safety: An approach for
20
developing resilience early warning
indicators,” Safety Science, vol. 118, pp.
316–331, Oct. 2019.
[33] R. Benabou and J. Tirole, “Intrinsic and
Extrinsic Motivation,” Review of Economic
Studies, vol. 70, no. 3, pp. 489–520, Jul.
2003.
[34] F. Guay, R. J. Vallerand, and C. Blanchard,
“On the Assessment of Situational Intrinsic
and Extrinsic Motivation: The Situational
Motivation Scale (SIMS),” Motivation and
Emotion, vol. 24, pp. 175–213, 2000.
[35] B. M. Wiechman and S. T. Gurland, “What
happens during the free-choice period?
Evidence of a polarizing effect of extrinsic
rewards on intrinsic motivation,” Journal of
Research in Personality, vol. 43, no. 4, pp.
716–719, Aug. 2009.
[36] T. Franke, C. Attig, and D. Wessel, “A
Personal Resource for Technology
Interaction: Development and Validation of
the Affinity for Technology Interaction
(ATI) Scale,” International Journal of
Human-Computer Interaction, vol. 35, no. 6,
pp. 456–467, 2019.
21