Determining Potential Failures and Challenges in Data Driven
Endeavors: A Real World Case Study Analysis
Daniel Staegemann, Matthias Volk, Tuan Vu, Sascha Bosse, Robert Häusler, Abdulrahman Nahhas,
Matthias Pohl and Klaus Turowski
Magdeburg Research and Competence Cluster Very Large Business Applications,
Faculty of Computer Science, Otto-von-Guericke University Magdeburg, Magdeburg, Germany
Keywords: Data Driven, Big Data, Case Study, Failures, Analysis, Categorization.
Abstract: The utilization of data in general and big data in particular offers large opportunities, but is at the same time
accompanied by a huge number of potential causes for failure. To avoid those pitfalls when realizing such
undertakings, at the beginning, it is necessary to develop an in-depth understanding of those causes. This
contribution analyses twelve real world case studies, from the big data and related domains, which were facing
issues. The causes for the experienced problems were extracted and thereupon categorized, facilitating the
understanding of practitioners and researchers that are engaged in the big data domain. Furthermore, potential
avenues for future research are highlighted.
1 INTRODUCTION
With the growing amount (Yin and Kaynak 2015;
Dobre and Xhafa 2014) and complexity (Yang et al.
2017) of data produced by humanity and the
increasing expectations regarding its utilization (Jin
et al. 2015), traditional technologies and approaches
are often overstrained. For this reason, big data
projects are becoming a promising solution for those
challenges. While the term big data itself has no
single, universally utilized explanation (Hartmann et
al. 2016), the definition provided by the National
Institute of Standards and Technology (NIST) is
widely accepted. According to that definition, big
data “consists of extensive datasets primarily in the
characteristics of volume, velocity, variety, and/or
variability that require a scalable architecture for
efficient storage, manipulation, and analysis” (NIST
2019). The application areas of those technologies are
multifarious. Examples comprise, but are not limited
to, the construction industry (Bilal et al. 2016),
procurement (Staegemann et al. 2019c), tourism
(Gajdošík 2019), urban transportation management
(Fiore et al. 2019), civil protection (Wu and Cui 2018)
and weather data analysis (Onal et al. 2017).
However, even though the potentials of big data are
manifold and high (Müller et al. 2018; Bughin 2016;
Maroufkhani et al. 2019; Alharthi et al. 2017), the
same applies for the challenges and risks (Alharthi et
al. 2017; Philip Chen and Zhang 2014; Staegemann
et al. 2019b; Wenzel and van Quaquebeke 2018;
Staegemann et al. 2019a; Volk et al. 2019). Hence, to
ensure the best results, it is necessary to have a deep
understanding of those challenges, to avoid potential
pitfalls, when implementing projects. For this reason,
the following research question will be explored in
the course of this work:
What are major factors contributing to failure in
big data endeavors?
To answer that research question, an analysis of
twelve real world cases from the big data or conjoined
domains, with negative outcomes, is conducted. The
focus goes beyond pure big data applications, since
failures are not as commonly publicized as success
stories, which limits the available resources. To
provide value, and offer comprehensive insights into
the domain, it however appears reasonable, to cover a
variety of settings and objectives. Furthermore, big
data projects are a subset of data driven endeavors
(Günther et al. 2017). Therefore, it appears
reasonable, to slightly extend the scope to also
include those. Hence, the work is structured as
follows. After providing an introduction, the second
section describes the respective cases. Afterward, the
analysis and its findings are presented. Finally, a
conclusion is drawn, which also includes the
contemplation of limitations and future perspectives.
Staegemann, D., Volk, M., Vu, T., Bosse, S., Häusler, R., Nahhas, A., Pohl, M. and Turowski, K.
Determining Potential Failures and Challenges in Data Driven Endeavors: A Real World Case Study Analysis.
DOI: 10.5220/0009792504530460
In Proceedings of the 5th International Conference on Internet of Things, Big Data and Security (IoTBDS 2020), pages 453-460
ISBN: 978-989-758-426-8
Copyright
c
 2020 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
453

2 CASE STUDIES
To increase the success rate of future big data
analytics projects, it is at first necessary to understand
the common factors leading to failures in big data
endeavors. For this purpose, the analysis of examples
from the past can provide valuable insights.
Therefore, in the following, twelve real world cases
will be analysed under this aspect. Due to the wide
range of possible data driven projects, the cases were
chosen from heterogeneous contexts, regarding
application area, intent, occurred issues, and financial
background, facilitating a better insight into the topic
as a whole (Khan and van Wynsberghe 2008).
2.1 Kmart
The retail company Kmart found itself in a price
competition with Walmart in the mid-to-late 1990s.
However, in contrast to its competitor, it neglected
technical advancements and therefore had no
functioning “just-in-time” inventory management. As
a result, customers were often faced with out of stock
items, resulting in poor shopping experiences (24/7
Wall St. 2012; Turner 2003). Furthermore, Kmart did
also not decide on a true identity and instead tried to
appeal to everyone. This approach, however, makes it
hard to define clear goals, which are a prerequisite for
a purposeful analysis and the formulation of strategies
(Marr 2017, 21 f.). In contrast, the competitors,
Walmart and Target, defined their own niche
(Leinwand and Mainardi 2010). While Kmart
eventually at least acknowledged the need for
technical support in planning and invested in a
business intelligence platform in 2003 (BusinessWire
2003), the damage had already been done. As a result,
a major decline in business success occurred, which
led to a merger with the department store chain Sears
in 2005 (24/7 Wall St. 2012; Egan 2015).
2.2 UK National Health Service
In 2002, the UK National Health Service launched a
project to achieve a top-down digitization in the
healthcare system in England, implementing new
technologies and IT systems. The idea was to
incorporate the plethora of different applications and
databases to create one integrated solution. While the
initial budget and duration were estimated to be £6.2
billion and ten years, the lackluster inclusion of health
care end-users, professionals and facilities, project
management issues, technical difficulties, and
changing specifications resulted in major problems.
In September 2011 it was dismantled after spending
about £9.8 billion without achieving the desired
results (Justinia 2017; Syal 2013; Hefford 2011).
2.3 Solid Gold Bomb T-shirt Company
To increase the variety of offered T-shirts, the startup
wrote a software that autonomously created designs
by combining the phrase “Keep Calm and”, which
was coined by the British as “Keep Calm and Carry
On” during the second world war, with another verb
and pronoun that were randomly taken from a
prepared list. Subsequently, those shirts were
automatically offered on Amazon, without further
review by the employees. Since the list with the
words that were to be used was not carefully curated
and checked in advance, this procedure resulted in
slogans like “Keep Calm and Rape Her” or “Keep
Calm and Hit Her”, leading to a public outrage and
the shutdown of the shop’s Amazon account, which
was also its main distribution channel, resulting in a
severe reduction of sales for the company (Pagliery
2013; McVeigh 2013).
2.4 Target
New parenthood is one of the occasions, where
retailers have the highest chance of changing
customer’s shopping behavior. Therefore, in 2002,
Target started developing a pregnancy-prediction
system to attract potential soon-to-be parents by
sending out corresponding coupons before the
competition even knew about the pregnancy. While
sales significantly increased, the specific coupons
also resulted in disclosing pregnancies to third
persons. For instance, a pregnant teenager’s father
was informed of the pregnancy due to the received
ads, which were promoting baby-related items. This
kind of unsolicited notification not only violates the
privacy of the expectant mother, but can potentially
also result in negative consequences for her. As a
result, Target started mixing customer-specific
promotions with randomly selected offers, reducing
the assignability of the advertisements with life
circumstances while still presenting a selection of
highly relevant items (Duhigg 2012; Albert 2015).
2.5 Pinterest
In contrast to Target’s contentwise correct but
intrusive analysis of its customer’s life circumstances,
in 2014 the social media company Pinterest falsely
congratulated a share of its community on getting
married. While weddings and the corresponding
accessories constitute a significant part of the
IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security
454

website’s presented content, many of the recipients
were not planning to get married or at least in a
relationship and in some cases, there was not even a
clear connection between the consumed contents and
marriage at all. As a result, some of the falsely
addressed users complained on twitter, attracting
attention by further media. Later on, Pinterest
apologized, stating that they wanted to reach people
who are interested in wedding-related content and the
wording, suggesting the imminence of an actual
wedding, was a mistake (Kosoff 2014; Roy 2014).
2.6 Google Flu Trends
Google Flu Trends (GFT) was a system that analysed
the frequency of flu-related search terms input into
google to predict flu outbreaks. To do so, about 50
million search terms were mapped to 1152 data points
(Ginsberg et al. 2009) which were identified as flu
indicators by the U.S. Centers for Disease Control
and Prevention. While there might be a connection
between those searches and actual cases of the flu, the
system completely missed the H1N1 flu and often
overestimated flu rates by high margins. Furthermore,
GFT lacked the transparency for outstanding persons
to properly evaluate the results, turning it into a black
box and preventing the submission of concepts for
possible improvements to the algorithms. In 2013,
GFT’s public website was shut down. However, the
concept and the data itself are still being used. For
instance, researchers continue working on models and
concepts to predict flue occurrences ahead of time
(Pappas 2014; Lazer et al. 2014; Comstock 2015;
Pervaiz et al. 2012).
2.7 Ferrari
In 2010 Fernando Alonso lost the almost certain
Formula 1 championship during the last race of the
season. To win the title by himself, independent from
his competitors’ performance, a fourth place would
have sufficed. As it is common, he was using a race
strategy that was chosen by his chief race strategist.
The process was supported by a decision support
system (DSS). By design, the DSS offered only two
options to pick. According to his team’s regulation,
the strategist was required to choose one of those,
which implies disregarding all other possibilities.
However, in contrast to the other races of the season,
the DSS could not build upon previous experiences
with the course, because it was its first appearance in
the circuit. Furthermore, as another limitation of the
system, insights from the ongoing race were not
included. In theory, the strategist could have ignored
that rule, ordered a different strategy and potentially
secured the title. Yet, he would have taken a personal
risk by violating the policy. In case of a failure and
with a prevailing blame-culture in the team, he would
have been in a hard-to-justify position. As a result, he
abided the rules and even though he chose the better
of the two proposed options, it was inappropriate for
the status of the race and the particularities of the
course, resulting in a seventh place, losing the almost
certain title (Aversa et al. 2018).
2.8 OfficeMax
In 2014, the office supply company OfficeMax sent
an advertisement mail to one of their off-and-on
customers. Even though, that procedure itself is not
noteworthy, the address printed on the envelope not
only included the name of the recipient, but also,
instead of the name of his business, the addition that
his daughter had died in a car crash. While the
information was correct, and his daughter, along with
her boyfriend, actually died in a car crash in the
previous year, the occurrence was disturbing for the
addressee and traumatizing for his wife.
Subsequently, the question arose, how and why the
company even had the information. Furthermore, this
highly delicate information was falsely labelled,
being the only reason for its revelation. In the
following, instead of apologizing, the contacted
representatives doubted the customer’s claims,
leading to the media being notified of the story. While
the company later on apologized and referred to a (not
disclosed) data broker as the culprit, stating that they
never ordered information that exceeded the
addresses itself and also announced they would
implement additional filters to flag inappropriate
information, the damage to the affected family was
already done (Hill 2014; Pearce 2014).
2.9 US Election Prediction 2016
With the 2016 US presidential election campaign
approaching its end, the final vote between Donald
Trump and Hillary Clinton, many media outlets tried
to predict its outcome. For example, the renowned
statistician Nate Silver prognosticated a 71.4 percent
(Silver 2016) chance of Clinton becoming the next
president. In doing so, he was even rather
conservative compared to other forecasts. The New
York Times estimated Clinton’s chances at 85
percent, Reuters gave her 90 percent, the Huffington
Post 98 percent and the Princeton Election
Consortium even 99 percent. While the timeframe to
be forecasted was only a few days and there were
Determining Potential Failures and Challenges in Data Driven Endeavors: A Real World Case Study Analysis
455

plenty of data from polls as well as sophisticated
algorithms for the analysis, the results differed
severely from each other and even more from the real
outcome. While it is hard to precisely determine the
cause, several factors were mentioned in the review.
Those comprise mistakes in the sampling of the
questioned population, dishonest answers due to
societal pressure, the dynamic and often emotional
nature of elections, the sometimes irrational actions
of humans. But also the fact, that the public opinion
was used as the only indicator of voting behavior
instead of utilizing additional sources like the number
and facial expressions of the attendance at rallies or
social media limited the significance of the obtained
results (Stone 2017).
2.10 Orca
Another example related to presidential elections,
which took place four years earlier and potentially
even influenced the outcome, is the mobile data
analytics platform Orca. As his opponent Barack
Obama (Scherer 2012), republican candidate Mitt
Romney intended to harness the power of big data for
the steering and coordination of his 2012 campaign.
The idea was to analyse what was happening at
polling stations and subsequently use that knowledge
to direct the efforts of campaign volunteers towards
potential Romney voters in contested states who had
not yet voted. While this general strategy is not new
and has been applied for many years, this time it was
supposed to be digitized, therefore allowing for a real
time analysis in contrast to the previously used
physical lists. However, the system itself, as well as
the corresponding communication, were flawed. The
designated users were not properly trained, the
according materials were not adequately distributed
and many of the volunteers could not even log into
the system, which also crashed repeatedly. This, in
turn, resulted in a huge waste of not only money for
the development of Orca, but also the manpower of
the volunteers who otherwise could have had a
positive impact on the outcome of the election, and
frustrated many of the most devoted supporters
(Casaretto 2012; Terkel 2012; Marcus 2012).
2.11 Facebook
With targeted advertisement being a major source of
income and over two billion users (Statista 2019), the
automated management of the ads constitutes a highly
important operation for Facebook. However, in 2017
it was discovered that the automatically created
categories for the definition of the target
demographic, which were based on the information
gathered on users, not only allowed explicitly
addressing “teachers” or “nurses”, but also a group of
about 2300 “Jew haters”. While this category, along
with others, has been subsequently deleted, a list of
around 5000 manually checked terms was curated and
a general increase in human oversight of its ad
targeting was promised by Facebook, this has not
been the first controversy regarding this part of the
company’s business (Elder 2014; Angwin and Parris
2016). Over the years, there were repeatedly issues
accompanying the automation and the options for
self-service and customization of the ads and their
distribution, leading to violations of the law, a loss of
trust and damage caused to the image. Yet, since this
automation is unavoidable, considering the
magnitude of the task, Facebook is challenged to
improve the underlying concepts and mechanisms
(Angwin et al. 2017; Dua 2017).
2.12 Tay
In 2016, Microsoft launched an AI chatbot named
Tay, which was supposed to interact with English
speaking people on Twitter, emulating the chatting
behavior of a teenage girl. The idea was to research
conversational understanding and at the same time
create interactions with a young target group. A
similar experiment had already been conducted in
China, where “XiaoIce” was running successfully.
However, this time, the endeavor did not proceed as
expected. After less than 24 hours and about 100.000
tweets, Tay had to be shut down, having turned into a
genocide promoting, anti-semitic racist. While many
of its insulting tweets were originated in a “repeat
after me” functionality and just copied messages that
were written by users, others were an effect of the AI
learning from the obtained inputs and various other
online sources. This resulted in statements like “bush
did 9/11 and Hitler would have done a better job than
the monkey we have now. donald trump is the only
hope we’ve got.” (Kleeman 2016). Besides shutting
Tay down, Microsoft also deleted the inappropriate
tweets and issued an apology, referring the scientific
and unpredictable nature of the experiment, but also
submitting an oversight on their part, underestimating
the potentially disruptive nature of the internet,
respectively a part of its users. This incident not only
showed how an internet-based-learning AI could be
corrupted, even by comparatively few people, and
therefore, how important a sufficient supervision is.
It also exemplified the enormous influence of cultural
differences in the usage of technologies and the
necessity
of taking those into account when creating
new concepts or implementing common ones
IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security
456

(Ohlheiser 2016; Vincent 2016; Steiner 2016; Hunt
2016).
3 ANALYSIS
This work mainly aims to deepen the understanding
regarding the practical implementation of big data in
enterprises by analyzing and categorizing the
mistakes in the presented cases. To increase the
objectivity in the extraction of those mistakes, three
of the authors have independently analyzed them.
Afterward the results were conjointly preprocessed.
This included, for example, the merging of synonyms
and the aggregation of similar causes. In cases where
divergences occurred, those were collectively
discussed to finally create a joint list of identified
failures (Orwin and Vevea 2009). The final list
comprises the factors Bad Data Quality, Ensuring
Data Security, Problematic Data Integration, Lack
Of Technological Understanding, Intransparent
Analysis, Relying Too Much On The System,
Insufficient Quality Assurance, Unrealistic
Budgeting, Wrong Data Interpretation, Insufficient
End-User Engagement, Lack Of Skills, Privacy
Concerns, Lack Of Vision, Bad Project Management,
Bad Company Culture and Lack Of Strategy.
Subsequently, using the same approach as before, the
identified failures were classified following the
categories presented in (Alharthi et al. 2017). Those
are Technology, Human, and Organization. While in
the first category, the factors are caused by technical
issues in a broader sense, and human causes can be
attributed to the actions and properties of individuals,
organizational issues originate from relations,
customs, structures and hierarchies within those
organizations. This also makes them probably the
hardest to overcome, since isolated measures are
unlikely to resolve the issue and extensive
interventions or restructurations have to be
performed. Furthermore, to take account of the
complexity of the regarded subject and the inherent
connections, also combinations of the three proposed
main categories were regarded. Those can involve
two or even all three of them, generating four more
classes, for a total of seven general fault types. The
results of the analysis, also constituting the answers
to the research question, are depicted in Figure 1,
showing the processed list of causes for failure as well
as their mapping to the categories. While the applied
procedure severely differs from the approach
followed in (Weibl and Hess 2018), the results show
large similarities, suggesting their validity.
4 CONCLUSION
Big data does not only offer opportunities, but is also
highly susceptible to failure. In the publication at
hand,
twelve real world cases, whose data driven
Figure 1: Factors contributing to the failure of data driven endeavors.
Determining Potential Failures and Challenges in Data Driven Endeavors: A Real World Case Study Analysis
457

approaches failed or had drawbacks of varied
severity, were analyzed and the determined issues
categorized, providing insights to practitioners and
researchers. In the future, the number of analyzed
cases shall be further increased and the results will be
fused with insights from existing studies as well as
expert interviews, to enhance the significance of the
findings. As a subsequent step, it is also intended to
use the enhanced data basis to develop concrete
solutions on how to avoid the identified pitfalls.
REFERENCES
24/7 Wall St. (2012): The Worst Business Decisions of All
Time. Available online at https://247wallst.com/
special-report/2012/10/17/the-worst-business-
decisions-of-all-time/, updated on 10/17/2012, checked
on 12/18/2019.
Albert, Kendra (2015): But What Did the Daughter Think?
Available online at https://medium.com/
@Kendra_Serra/but-what-did-the-daughter-think-
8d9233789b4f, updated on 8/31/2015, checked on
12/18/2019.
Alharthi, Abdulkhaliq; Krotov, Vlad; Bowman, Michael
(2017): Addressing barriers to big data. In Business
Horizons 60 (3), pp. 285–292.
Angwin, Julia; Parris, Terry (2016): Facebook Lets
Advertisers Exclude Users by Race. Available online at
https://www.propublica.org/article/facebook-lets-
advertisers-exclude-users-by-race, updated on
10/28/2016, checked on 1/3/2020.
Angwin, Julia; Varner, Madeleine; Tobin, Ariana (2017):
Facebook let advertisers target ‘Jew Haters’. Available
online at https://www.businessinsider.com/facebook-
let-advertisers-reach-jew-haters-through-ad-buying-
tool-2017-9?r=DE&IR=T, updated on 9/14/2017,
checked on 1/3/2020.
Aversa, Paolo; Cabantous, Laure; Haefliger, Stefan (2018):
When decision support systems fail: Insights for
strategic information systems from Formula 1. In The
Journal of Strategic Information Systems 27 (3),
pp. 221–236.
Bilal, Muhammad; Oyedele, Lukumon O.; Qadir, Junaid;
Munir, Kamran; Ajayi, Saheed O.; Akinade, Olugbenga
O. et al. (2016): Big Data in the construction industry:
A review of present status, opportunities, and future
trends. In Advanced Engineering Informatics 30 (3),
pp. 500–521.
Bughin, Jacques (2016): Big data, Big bang? In Journal of
Big Data 3 (1).
BusinessWire (2003): Kmart Selects Business Objects as
Its Business Intelligence Platform. Available online at
https://www.businesswire.com/news/home/200306100
05293/en/Kmart-Selects-Business-Objects-Business-
Intelligence-Platform, updated on 6/10/2003, checked
on 12/18/2019.
Casaretto, John (2012): Romney’s Project Orca – a Big
Data Fail. Available online at https://siliconangle.com/
2012/11/12/romneys-project-orca-a-big-data-fail/,
updated on 11/12/2012, checked on 1/1/2020.
Comstock, Jonah (2015): Google Flu Trends website shuts
down; will send data to Boston Children's, Columbia,
CDC. Available online at https://
www.mobihealthnews.com/46248/google-flu-trends-
website-shuts-down-will-send-data-to-boston-
childrens-columbia-cdc, updated on 8/21/2015,
checked on 12/18/2019.
Dobre, Ciprian; Xhafa, Fatos (2014): Intelligent services
for Big Data science. In Future Generation Computer
Systems 37, pp. 267–281.
Dua, Tanya (2017): Facebook promises more human
oversight of its ad targeting, as COO Sheryl Sandberg
says recent anti-Semitic mishap is a 'fail on our part'.
Available online at https://www.businessinsider.com/
facebook-sheryl-sandberg-introduces-manual-reviews-
of-ad-targeting-after-propublica-article-reveals-anti-
semitic-categories-2017-9?r=DE&IR=T, updated on
9/20/2017, checked on 1/3/2020.
Duhigg, Charles (2012): How Companies Learn Your
Secrets. Available online at https://www.nytimes.com/
2012/02/19/magazine/shopping-habits.html?
ref=magazine, updated on 2/16/2012, checked on
12/18/2019.
Egan, Matt (2015): Kmart's sales have fallen off a gigantic
cliff. Available online at https://money.cnn.com/
2015/06/08/investing/kmart-sales-decline-sears-eddie-
lampert/, updated on 6/9/2015, checked on 12/18/2019.
Elder, Jeff (2014): Nude Webcams and Diet Drugs: the
Facebook Ads Teens Aren't Supposed to See. Available
online at https://www.wsj.com/articles/facebook-ads-
teens-werent-supposed-to-see-1393541465, updated on
2/27/2014, checked on 1/3/2020.
Fiore, Sandro; Elia, Donatello; Pires, Carlos Eduardo;
Mestre, Demetrio Gomes; Cappiello, Cinzia; Vitali,
Monica et al. (2019): An Integrated Big and Fast Data
Analytics Platform for Smart Urban Transportation
Management. In IEEE Access 7, pp. 117652–117677.
Gajdošík, Tomáš (2019): Big Data Analytics in Smart
Tourism Destinations. A New Tool for Destination
Management Organizations? In Vicky Katsoni, Marival
Segarra-Oña (Eds.)Smart Tourism as a Driver for
Culture and Sustainability, vol. 77. Cham: Springer
International Publishing (Springer Proceedings in
Business and Economics), pp. 15–33.
Ginsberg, Jeremy; Mohebbi, Matthew H.; Patel, Rajan S.;
Brammer, Lynnette; Smolinski, Mark S.; Brilliant,
Larry (2009): Detecting influenza epidemics using
search engine query data. In Nature 457, 1012 - 1014.
Günther, Wendy Arianne; Rezazade Mehrizi, Mohammad
Hosein.; Huysman, Marleen; Feldberg, Frans (2017):
Debating big data: A literature review on realizing
value from big data. In The Journal of Strategic
Information Systems 26 (3), pp. 191–209.
Hartmann, Philipp Max; Zaki, Mohamed; Feldmann, Niels;
Neely, Andy (2016): Capturing value from big data – a
taxonomy of data-driven business models used by start-
IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security
458

up firms. In International Journal of Operations &
Production Management 36 (10), pp. 1382–1406.
Hefford, Rhys (2011): Why the NHS National Programme
for IT didn't work. Available online at https://www.cio.
co.uk/it-strategy/why-nhs-national-programme-for-it-
didnt-work-3431723/, updated on 12/2/2011, checked
on 12/18/2019.
Hill, Kashmir (2014): OfficeMax Blames Data Broker For
'Daughter Killed in Car Crash' Letter. Available online
at
https://www.forbes.com/sites/kashmirhill/2014/01/22/
officemax-blames-data-broker-for-daughter-killed-in-
car-crash-letter, updated on 1/22/2014, checked on
1/2/2020.
Hunt, Elle (2016): Tay, Microsoft's AI chatbot, gets a crash
course in racism from Twitter. Available online at
https://www.theguardian.com/technology/2016/mar/24
/tay-microsofts-ai-chatbot-gets-a-crash-course-in-
racism-from-twitter, updated on 3/24/2016, checked on
1/4/2020.
Jin, Xiaolong; Wah, Benjamin W.; Cheng, Xueqi; Wang,
Yuanzhuo (2015): Significance and Challenges of Big
Data Research. In Big Data Research 2 (2), pp. 59–64.
Justinia, Taghreed (2017): The UK's National Programme
for IT: Why was it dismantled? In Health services
management research 30 (1), pp. 2–9.
Khan, Samia; van Wynsberghe, Robert (2008): Cultivating
the Under-Mined. Cross-Case Analysis as Knowledge
Mobilization. In Forum: Qualitative Social Research 9
(1).
Kleeman, Sophie (2016): Here Are the Microsoft Twitter
Bot’s Craziest Racist Rants. Available online at
https://gizmodo.com/here-are-the-microsoft-twitter-
bot-s-craziest-racist-ra-1766820160, updated on
3/24/2016, checked on 1/4/2020.
Kosoff, Maya (2014): Pinterest Accidentally Sent Emails
To Single Women Congratulating Them On Getting
Married. Available online at https://
www.businessinsider.com/pinterest-accidental-
marriage-emails-2014-9?r=DE&IR=T, updated on
9/4/2014, checked on 1/1/2020.
Lazer, David; Kennedy, Ryan; King, Gary; Vespignani,
Alessandro (2014): Big data. The parable of Google
Flu: traps in big data analysis. In Science (New York,
N.Y.) 343 (6176), pp. 1203–1205.
Leinwand, Paul; Mainardi, Cesare (2010): Why Can’t
Kmart Be Successful While Target and Walmart
Thrive? Available online at https://hbr.org/
2010/12/why-cant-kmart-be-successful-w, updated on
12/15/2010, checked on 12/18/2019.
Marcus, Stephanie (2012): Mitt Romney’s Project ORCA
Failure: Broken ORCA App Cost Him Thousands Of
Votes. Available online at https://www.huffpost.com/
entry/mitt-romney-project-orca-broken-app-cost-
thousands-votes_n_2109986, updated on 11/10/2012,
checked on 1/1/2020.
Maroufkhani, Parisa; Wagner, Ralf; Wan Ismail, Wan
Khairuzzaman; Baroto, Mas Bambang; Nourani,
Mohammad (2019): Big Data Analytics and Firm
Performance: A Systematic Review. In Information 10
(7), p. 226.
Marr, Bernard (2017): Data strategy. How to profit from a
world of big data, analytics and the internet of things.
New York: Kogan Page Ltd. Available online at
http://search.ebscohost.com/login.aspx?direct=true&sc
ope=site&db=nlebk&AN=1494509.
McVeigh, Tracy (2013): Amazon acts to halt sales of 'Keep
Calm and Rape' T-shirts. Available online at
https://www.theguardian.com/technology/2013/mar/02
/amazon-withdraws-rape-slogan-shirt, updated on
3/2/2013, checked on 12/18/2019.
Müller, Oliver; Fay, Maria; Vom Brocke, Jan (2018): The
Effect of Big Data and Analytics on Firm Performance:
An Econometric Analysis Considering Industry
Characteristics. In Journal of Management Information
Systems 35 (2), pp. 488–509.
NIST (2019): NIST Big Data Interoperability Framework:
Volume 1, Definitions, Version 3. Gaithersburg, MD.
Ohlheiser, Abby (2016): Trolls turned Tay, Microsoft’s fun
millennial AI bot, into a genocidal maniac. Available
online at https://www.washingtonpost.com/news/the-
intersect/wp/2016/03/24/the-internet-turned-tay-
microsofts-fun-millennial-ai-bot-into-a-genocidal-
maniac/, updated on 3/25/2016, checked on 1/4/2020.
Onal, Aras Can; Berat Sezer, Omer; Ozbayoglu, Murat;
Dogdu, Erdogan (2017): Weather data analysis and
sensor fault detection using an extended IoT framework
with semantics, big data, and machine learning. In Jian-
Yun Nie, Zoran Obradovic, Toyotaro Suzumura, Rumi
Ghosh, Raghunath Nambiar, Chonggang Wang
(Eds.)Proceedings of the 2017 IEEE International
Conference on Big Data. Boston, MA, 11.12.2017-
14.12.2017. Piscataway, NJ: IEEE, pp. 2037–2046.
Orwin, Robert G.; Vevea, Jack L. (2009): Evaluating
Coding Decisions. In Harris M. Cooper, Larry V.
Hedges, Jeff C. Valentine (Eds.)The handbook of
research synthesis and meta-analysis. 2nd ed. New
York: Russell Sage Foundation, pp. 177–206.
Pagliery, Jose (2013): Man behind 'Carry On' T-shirts says
company is 'dead'. Available online at https://money.
cnn.com/2013/03/05/smallbusiness/keep-calm-and-
carry-on/, updated on 3/5/2013, checked on 12/18/2019.
Pappas, Stephanie (2014): Data Fail! How Google Flu
Trends Fell Way Short. Available online at
https://www.livescience.com/44089-google-flu-trends-
problems.html, updated on 3/13/2014, checked on
12/18/2019.
Pearce, Matt (2014): OfficeMax executive apologizes over
‘daughter killed’ mailer. Available online at
https://www.latimes.com/nation/la-na-officemax-
mess-20140121-story.html, updated on 1/20/2014,
checked on 1/2/2020.
Pervaiz, Fahad; Pervaiz, Mansoor; Abdur Rehman, Nabeel;
Saif, Umar (2012): FluBreaks: early epidemic detection
from Google flu trends. In Journal of medical Internet
research 14 (5), e125.
Philip Chen, C. L.; Zhang, Chun-Yang (2014): Data-
intensive applications, challenges, techniques and
Determining Potential Failures and Challenges in Data Driven Endeavors: A Real World Case Study Analysis
459

technologies: A survey on Big Data. In Information
Sciences 275, pp. 314–347.
Roy, Jessica (2014): Pinterest Accidentally Congratulates
Single Women on Getting Married. Available online at
http://nymag.com/intelligencer/2014/09/pinterest-
congratulates-single-women-on-marriage.html?,
updated on 9/4/2014, checked on 1/1/2020.
Scherer, Michael (2012): Inside the Secret World of the
Data Crunchers Who Helped Obama Win. Available
online at http://swampland.time.com/2012/11/07/
inside-the-secret-world-of-quants-and-data-crunchers-
who-helped-obama-win/, updated on 11/7/2012,
checked on 1/1/2020.
Silver, Nate (2016): Final Election Update: There’s A Wide
Range Of Outcomes, And Most Of Them Come Up
Clinton. Available online at https://fivethirtyeight.com/
features/final-election-update-theres-a-wide-range-of-
outcomes-and-most-of-them-come-up-clinton/,
updated on 11/8/2016, checked on 12/31/2019.
Staegemann, Daniel; Hintsch, Johannes; Turowski, Klaus
(2019a): Testing in Big Data: An Architecture Pattern
for a Development Environment for Innovative,
Integrated and Robust Applications. In Proceedings of
the WI2019, pp. 279–284.
Staegemann, Daniel; Volk, Matthias; Jamous, Naoum;
Turowski, Klaus (2019b): Understanding Issues in Big
Data Applications – A Multidimensional Endeavor. In
25th Americas Conference on Information Systems,
AMCIS 2019, Cancun, Q.R, Mexico, August 15-17,
2019: Association for Information Systems.
Staegemann, Daniel; Volk, Matthias; Turowski, Klaus
(2019c): Mobile Procurement Management. In Tobias
Kollmann (Ed.)Handbuch Digitale Wirtschaft, vol. 9.
Wiesbaden: Springer Fachmedien Wiesbaden (Springer
Reference Wirtschaft), pp. 1–15.
Statista (2019): Number of monthly active Facebook users
worldwide as of 3rd quarter 2019. Available online at
https://www.statista.com/statistics/264810/number-of-
monthly-active-facebook-users-worldwide/, updated
on 11/19/2019, checked on 1/3/2020.
Steiner, Anna (2016): Was Microsoft durch „Tay“ gelernt
hat. Available online at https://www.faz.net/
aktuell/wirtschaft/netzwirtschaft/was-microsoft-mit-
dem-bot-tay-von-der-netzgemeinde-gelernt-hat-
14146188.html, updated on 3/26/2016, checked on
1/4/2020.
Stone, Adam (2017): When Big Data Gets It Wrong.
Available online at https://www.govtech.com/data/
When-Big-Data-Gets-It-Wrong.html, updated on
March 2017, checked on 12/31/2019.
Syal, Rajeev (2013): Abandoned NHS IT system has cost
£10bn so far. Available online at https://
www.theguardian.com/society/2013/sep/18/nhs-
records-system-10bn, updated on 9/18/2013, checked
on 12/18/2019.
Terkel, Amanda (2012): Project ORCA: Mitt Romney
Campaign Plans Massive, State-Of-The-Art Poll
Monitoring Effort. Available online at https://
www.huffpost.com/entry/project-orca-mitt-
romney_n_2052861, updated on 12/6/2017, checked on
1/1/2020.
Turner, Marcia Layton (2003): Kmart's ten deadly sins.
How rogue managers ruined an American icon. New
York, Chichester: Wiley.
Vincent, James (2016): Twitter taught Microsoft’s AI
chatbot to be a racist asshole in less than a day.
Available online at https://www.theverge.com/
2016/3/24/11297050/tay-microsoft-chatbot-racist,
updated on 3/24/2016, checked on 1/4/2020.
Volk, Matthias; Staegemann, Daniel; Pohl, Matthias;
Turowski, Klaus (2019): Challenging Big Data
Engineering: Positioning of Current and Future
Development. In Proceedings of the 4th International
Conference on Internet of Things, Big Data and
Security. 4th International Conference on Internet of
Things, Big Data and Security. Heraklion, Crete,
Greece, 02.05.2019 - 04.05.2019: SCITEPRESS -
Science and Technology Publications, pp. 351–358.
Weibl, Johannes; Hess, Thomas (2018): Success or Failure
of Big Data: Insights of Managerial Challenges from a
Technology Assimilation Perspective. In Proceedings
of the MKWI 2018. Multikonferenz
Wirtschaftsinformatik 2018. Lüneburg, Germany,
06.03.2018-09.03.2018, pp. 47–58.
Wenzel, Ramon; van Quaquebeke, Niels (2018): The
Double-Edged Sword of Big Data in Organizational
and Management Research. In Organizational
Research Methods 21 (3), pp. 548–591.
Wu, Desheng; Cui, Yiwen (2018): Disaster early warning
and damage assessment analysis using social media
data and geo-location information. In Decision Support
Systems 111, pp. 48–59.
Yang, Chaowei; Huang, Qunying; Li, Zhenlong; Liu, Kai;
Hu, Fei (2017): Big Data and cloud computing:
innovation opportunities and challenges. In
International Journal of Digital Earth 10 (1), pp. 13–
53.
Yin, Shen; Kaynak, Okyay (2015): Big Data for Modern
Industry: Challenges and Trends [Point of View]. In
Proceedings of the IEEE 103 (2), pp. 143–146.
IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security
460