Examining Competing Entrepreneurial Concerns in a Social
Question and Answer (SQA) Platform
Irawan Nurhas
1,2 a
, Henri Pirkkalainen
3b
, Stefan Geisler
1c
and Jan Pawlowski
1,2 d
1
Institute of Positive Computing, Hochschule Ruhr West University of Applied Sciences, Bottrop, Germany
2
Faculty of Information Technology, University of Jyväskylä, Jyväskylä, Finland
3
Faculty of Management and Business, Tampere University, Tampere, Finland
Keywords: Entrepreneurial Concerns, Challenges to Startups, Topic Modeling, Entrepreneurship, Sentiment Analysis.
Abstract: This study aims to determine the competing concerns of people interested in startup development and
entrepreneurship by using topic modeling and sentiment analysis on a social question-and-answer (SQA)
website. Understanding the underlying concerns of startup entrepreneurs is critical to society and economic
growth. Therefore, greater scientific support for entrepreneurship remains necessary, including data mining
from virtual social communities. In this study, an SQA platform was used to identify the sentiment of thirty
concerns of people interested in startup entrepreneurship. Based on topic modeling and sentiment analysis of
18819 inquiries in various forums on an SQA, we identified additional questions about founder figures, keys
to success, and the location of a startup. In addition, we found that general questions were rated more
positively, especially when it came to pitching, finding good sources, disruptive innovation, idea generation,
and marketing advice. On average, the identified concerns were considered 48.9 percent positive, 41 percent
neutral, and 10.1 percent negative. This research establishes a critical foundation for future research and
development of digital startups by outlining a variety of different concerns associated with startup
development in the digital age.
1 INTRODUCTION
This study provides a comprehensive insight into
people's concerns interested in entrepreneurship and
startup entrepreneurs (SEs). Entrepreneurship is a
powerful economic and social force (Harb & Shang,
2021; Schöning, 2013; van Stel et al., 2005). As a
result, several initiatives have been launched to
encourage SEs to establish their enterprises or
business models (Ratinho et al., 2020).
Meanwhile, in the knowledge economy and age
of virtual communities, the Social Question-and-
Answer (SQA) platform is rapidly moving
information-seeking behavior toward a more
collaborative and personalized question-and-answer
experience based on expertise (Choi et al., 2014).
Thus, SQA provides access to business networking
and open knowledge- and experience-based business
a
https://orcid.org/0000-0002-2211-8857
b
https://orcid.org/0000-0002-5389-7363
c
https://orcid.org/0000-0002-1976-0013
d
https://orcid.org/0000-0002-7711-1169
and entrepreneurial solutions to SEs with minimal
resources (Shneor & Flåten, 2015).
Although numerous studies on SEs have been
conducted (Chandra et al., 2016; Puhakka & Ojala,
2021; Saura et al., 2019), it is important to determine
the concerns of the discussed topic on a large scale
to capture better the different and more detailed
aspects (Saura et al., 2019). Therefore, topic
modeling (Chandra et al., 2016; Onan et al., 2016;
Rehurek & Sojka, 2010) and sentiment analysis
(Gao et al., 2013; Hutto & Gilbert, 2014; Onan et al.,
2016) methods were applied to the questions
collected from an SQA.
By investigating the substance of the issues in an
SQA, we set out to identify which competing
concerns are most salient and how these concerns are
perceived and evolve over time in the face of
digitalization. This study contributes twofold: First,
Nurhas, I., Pirkkalainen, H., Geisler, S. and Pawlowski, J.
Examining Competing Entrepreneurial Concerns in a Social Question and Answer (SQA) Platform.
DOI: 10.5220/0010661000003064
In Proceedings of the 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2021) - Volume 3: KMIS, pages 145-152
ISBN: 978-989-758-533-3; ISSN: 2184-3228
Copyright
c
2021 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
145
we identify thirty areas of concern that emphasize the
most disputed issues involving positive and negative
emotions, thus paving the groundwork for future
scholarly investigation. Second, this study expands
the body of knowledge based on previous research by
emphasizing the use of SQA as a valuable element for
researching entrepreneurship and confirming the
identified SEs' dilemmas through the lens of SQA,
spanning a broader range of SE development topics
that are not limited by cultural, country-specific,
gender, or age boundaries.
As a reminder, the research presentation should
adhere to the following structure. In the next section,
we present studies on entrepreneurial challenges for
startups and the use of data mining techniques to
identify entrepreneurial insights. Subsequently, we
describe how we conducted our research using topic
modeling and sentiment analysis. We then highlight
the competing concerns of SEs in our findings
section. Next, we address the study's findings,
limitations, and recommendations for future research
in the following discussion section-and; finally, we
provide an overview of the study report.
2 RELATED WORKS
SEs face a variety of challenges in the early stages or
throughout their entrepreneurial journey (Giardino et
al., 2015; Nurhas et al., 2020; Saura et al., 2019;
Wang et al., 2016), including issues related to product
and team development, lean process, business model
design, and financing, scaling, partnerships, sales,
and product and market alignment (Giardino et al.,
2015; Wang et al., 2016). These challenges motivate
SEs to freely discuss, share, and communicate
attached with emotion in order to highlight the
elements of success, concerns, or motivations (Saura
et al., 2019) that spread across the Internet, the social
media community, including SQA (Chandra et al.,
2016; Puhakka & Ojala, 2021; Shneor & Flåten,
2015).
Giardino et al. (2015) highlighted that in addition
to the team and product-related challenges,
technology uncertainties and acquiring the first
customer are prominent challenges for software
startups that are also relevant to developing a
technology-based startup in the digital age. However,
in the study based on the analysis of entrepreneurial
posts on social media, these issues were not
mentioned as essential concerns for success factors
for SEs (Saura et al., 2019).
1
https://www.twitter.com/
The potential of identifying the topic of concerns
of SEs has already been investigated in a study of
shared thoughts on Twitter
1
, where topics were
categorized into business angels, business plans,
methodology, tools, projects, jobs, and founders
(Saura et al., 2019). Topic modeling and sentiment
analysis are two well-known tools (Gao et al., 2013;
Onan et al., 2016) for revealing the categories of
concern and emotions behind shared thoughts and
have also been used to study entrepreneurship and
startup based on a large amount of data collected
(Chandra et al., 2016; Harb & Shang, 2021; Puhakka
& Ojala, 2021; Saura et al., 2019). The methods have
already been used in various fields and SQA (Jiang et
al., 2018; Kumar et al., 2018).
Uncovering topics and sentiments from SQA is
important to identify motivations, areas of interest,
and expectations that will raise community awareness
of specific issues (Choi et al., 2014). Furthermore,
sentiment analysis has been demonstrated to reveal
feelings and perspectives that can influence decision-
making on particular concerns (Chen et al., 2018;
Choi et al., 2014; Johnson, 1990). Consequently, it is
critical for triggering discussion, directing, and
promoting the selection of policies, strategies, and
approaches to entrepreneurial development
(Blanchflower & Oswald, 1998; Gifford, 1992;
Nurhas et al., 2020; O’Shea et al., 2017).
3 METHODS
Two commonly accepted strategies for doing SQA
research are user-based and content-based research
(Choi et al., 2014). This research involved the use of
a content-based SQA method that focuses on content
patterns rather than user characteristics (Choi et al.,
2014). The Selenium Python package was used to
extract snapshot data at the beginning of the process
(Huber et al., 2011). The data was compiled from the
text of questions posted in the SQA Quora
2
forums
dedicated to startups and entrepreneurship.
We chose Quora as our study platform because it
appears to be a promising source for investigating
various social phenomena in other studies (Jiang et
al., 2018; Wang et al., 2013). Quora is an online
community founded in 2009 and headquartered in
California, United States of America. Quora focuses
on an online open-knowledge that is based on user-
generated content. Quora's primary content is a
question posed by a user to which other users
respond. Users of Quora are expected to use their real
2
https://www.quora.com/
KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems
146
names, though they can choose to remain anonymous.
Quora is ranked #356 globally by AlexaPage Rank
(as of June 15, 2021) and is available in many
countries. Even though Quora's content is open, the
company does not offer an official API for retrieving
data from its websites.
In the following instance, data collection was
done by taking all the data in Quoran SQA on
September 21, 2018, removing duplicate questions
(N_initial:24802 questions to N_final: 18819
questions), cleaning the data by removing
punctuations and stop words, developing labels for
different topics as concerns, and identifying the
sentiment of each topic to identify competing
concerns. The Latent Dirichlet Allocation (LDA)
algorithm of Gensim was used as a Python package
for clustering topics because it is considered to
outperform other known topic models (Harb &
Shang, 2021).
The LDA class of statistical language models is
used in generative probability computing, which is a
subset of statistical computing. LDA makes Topics
using word clusters instead of text clusters to
understand data better (specifically, in this study, the
SQA's questions). In LDA, each topic is a model of a
mixture of words. Each word is represented by a
Dirichlet distribution coefficient. The model in each
topic has the capability to forecast significantly
related topics in a document (Blei et al., 2003). For
LDA, a predictive likelihood-based approach was
utilized to pick out the most optimal number of topics
(Chang et al., 2009). Then, the concern was labeled
based on the given keywords and the coefficient of
each identified topic from the Gensim LDA model.
Three to ten relevant keywords of word clusters were
used to designate each topic.
Following the identification of each cluster's
pertinent topics, the next step is to ascertain the
sentiment for each question posed within each cluster.
The Vader algorithm (Hutto & Gilbert, 2014) was
used to identify sentiment for each issue, which
shows not only sentiments and viewpoints from
written language but also emojis or spoken emotions
(Hutto & Gilbert, 2014; Onan et al., 2016) that are
relevant and frequently contained in a text (Gao et al.,
2013). All queries that have the same identified topic
were grouped individually. The Vader algorithm
determines each question's negative, positive, and
neutral trends on the same topic. The percentage of
sentiments for each issue was derived by comparing
the three tendencies to the total number of questions
on a specific topic of concern.
Following that, each topic's popularity ratio was
determined and organized in descending order
(1 to 30, where 30 is the most popular topic in a
given year; and 0 for topics not found in a given
year, this may happen, for example, if a topic
appears after 2010). The popularity ratio value is the
percentage of questions asked about a specific topic
in a given year compared to the total number of
inquiries from all topics. It is compared to the
percentage of each topic to detect dynamic changes
in topic popularity.
4 RESULT
Based on the log-likelihood value of various learning
decay scenarios (0.5, 0.7, & 0.9). Thirty topics with a
consistent value amongst diverse scenarios of
concern were taken into account. As shown in Figure
1, thirty topics were chosen as the number that can
yield the greatest number of topics for the dataset.
The selection is based on the maximum feasible log-
likelihood value under specific conditions before the
log-likelihood value for a larger number of topics
drops significantly.
Figure 1: log-likelihood value for a different number of
topics.
Table 1 presents the thirty labeled concerns as a
result of defining the cluster from LDA. This
includes, for example, lean startup, the
entrepreneurial process, pitching the idea, scaling the
business, and managing company resources, such as
founders and co-founders, investments, and location,
partnerships, and emerging trends (Giardino et al.,
2015; Ratinho et al., 2020; Shneor & Flåten, 2015;
Wang et al., 2016).
Overall, sentiment for each concern was positive,
with an average percentage of 48.9 percent. All
competing concerns for positive and negative were
outlined in gray in Table 1. For example, concerns
about Pitching and finding good sources with 66%
and concern about success keys (64.8%) were
Examining Competing Entrepreneurial Concerns in a Social Question and Answer (SQA) Platform
147
Table 1: List of concerns of SE and the identified sentiment.
Label of
concern
Top 3 words and the
coefficient
N
Percentage of
sentiment (%)
(-) (n) (+)
Lean
startup
0.34*"startup"+
0.18*"lean"+
0.05*"book"
751 7.1 53.5 39.4
Inspiring
Company
0.11*"company"+
0.12*"grow"+
0.06*"learn"
743 10.2 42.8 47.0
Founder
figure
0.17*"elon_musk"+
0.05*"time"+0.04*"t
esla"
895
24.4 40.2 35.4
Find co-
founder
0.18*"startup"+
0.11*"founder"+0.0
5*"hire"
756 10.7 44.3 45.0
Success
keys
0.19*"entrepreneur"
+
0.15*"successful"+
0.03*"key"
881 10.4 24.7 64.8
User-
centric
app
0.12*"build"+
0.08*"app"+0.06*"u
ser"'
658 5.6 43.6 50.8
A year of
remote
wor
k
0.18*"work"+
0.07*"year"+0.06*"j
ob"
701
17.1 43.5 39.4
Payment
service
0.13*"service" +
0.06*"pay"+0.05*"p
roblem"
672
16.4 32.9 50.7
Pitching
0.12*"innovative"+
0.09*"idea"+0.06*"s
ell"
653 6.9 27.1 66.0
Investmen
t capital
0.27*"start"+
0.11*"business"+
0.07*"investment"
623 10.1 49.9 40.0
Scaling
the
b
usiness
0.21*"small"+
0.19*"business"+
0.06*"scale"
521 8.3 49.7 42.0
Breakeven
point
0.25*"make"+
0.14*"people"+
0.11*"mone
y
"
716
15.6 42.0 42.3
Small
business
owne
r
0.22*"business"+
0.18*"small"+
0.08*"owner"
540 11.7 45.4 43.0
Disruptive
innovation
0.18*"innovation"+
0.04*"change"+
0.03*"life"
787 8.8 30.9
60.4
Important
items
0.09*"thing"+
0.07*"important" +
0.06*"open"
647 7.4 41.7 50.9
Find good
sources
0.28*"good"+
0.14*"find"+
0.07*"software"
447 5.8 28.2
66.0
Funding
0.20*"startup"+
0.06*"investor"+
0.05*"invest"
608 9.7 46.4 43.9
Startup
location
0.32*"startup"+
0.32*"tech"+
836 8.4 45.2 46.4
Label of
concern
Top 3 words and the
coefficient
N
Percentage of
sentiment
(
%
)
(
-
)
(
n
)
(
+
)
0.02*"silicon_valley
"
Product
launch
0.25*"product"+
0.11*"market"+
0.08*"launch"
672 8.2 38.5 53.3
Initate
new
b
usiness
0.36*"business"+
0.08*"run" +
0.08*"start"
513 5.3 35.5 59.3
Risk
manageme
nt
0.27*"strategy"+
0.04*"fail"+
0.03*"corporate"
684 13.0 45.2 41.8
Business
plan
0.40*"business"+
0.15*"plan"+
0.12*"good"
486 5.8 50.6 43.6
Business
model
0.12*"business" +
0.10*"model"+
0.08*"strategic"
591 6.4 50.3 43.3
Marketing
advise
0.08*"marketing"+
0.07*"give"+
0.07*"website"
531 7.3 32.4
60.3
Validating
idea
0.12*"customer"+
0.08*"development"
+ 0.05*"idea"
497 10.1 40.2 49.7
Digital
business
0.31*"business"+
0.25*"start"+
0.10*"online"
431 7.7 45.0 47.3
Emerging
markets
0.16*"start"+
0.13*"business"+
0.10*"india"
496 9.3 56.7 34.1
Partnershi
p
0.37*"company"+
0.04*"partner"+
0.03*" competitive_
advantage"
447
13.9 36.7 49.4
Technolog
y trends
0.11*"technology"+
0.09*"big"+
0.08*"industry"
509 9.4 41.3 49.3
Idea
invention
0.15*"idea"+
0.11*"invention"+
0.10*"good"
527 12.0 24.7
63.4
N: total questions; (-): negative sentiment; (+): positive
sentiment; (n): neutral sentiment;
positively perceived. Conversely, some issues are
perceived as challenges to SEs development, with
concerns about founder figure (24.4%), remote work
(17.1%), payment services, and break events (at
around 16% each) being perceived as more negative.
Based on Table 1 and in light of previous literature on
the dimensions of startup challenges, including
product, financial, team, and market (Giardino et al.,
2015; Ratinho et al., 2020; Shneor & Flåten, 2015;
Wang et al., 2016), we identified the following as
additional concerns of SEs that need to be considered.
As illustrated in Table 2, international startups and
role models are two additional dimensions of
KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems
148
challenges that have been overlooked previously and
may arise due to Quora's global nature. As a result,
people from various countries use the platform to
inquire about opportunities, challenges, and success
factors associated with developing startups in other
locations and require an example of successful
development.
The "role models" category is an example of how
digital technology can help small and medium-sized
enterprises (SMEs) find inspiring stories of role
models who are either individuals or organizations to
gain insights into how startups in situations similar to
their own are evolving.
Table 2: Further concern dimension for SEs.
(Challenge
category):
definition
Label of
concern
Example question from
Quora
(internationa
l startup):
lack of
knowledge
regarding
place
developing
international
startup
Startup
location
“Why has Sub-Saharan
Africa failed to produce tech
giants like Twitter,
Facebook, Apple, Google,
Arm (acquired by SoftBank),
etc.?”
“Is Silicon Valley really the
best tech startu
p
location?”
Emerging
market
“What are the ways to start
a tech startup in India while
bein
g
located in the USA?
“
(Role
model): lack
of inspiring
role model
Founder
Figure
“What Elon Musk Gets
Wrong About Leadership?”
“Why did Steve Jobs push
technology forward, even
risking his company to
fail?”
“Why are there relatively
few female tech startup
f
ounders?”
Inspiring
company
“I’m 16, and I want to start
my own software company
like Apple and Microsoft.
How and where should I
start with?”
“What do you learn in a
small company that you can't
learn in a bi
g
one?”
SQA provides a place for direct engagement with
other successful SEs when it comes to using SQA to
look for role models. They may have a comparable
background and can serve as inspiring role models.
As seen in Figure 2, the number of queries for all
topics climbed dramatically year after year. Based on
the percentage of the ratio value of each year's
calculation. We discovered some intriguing patterns
of concerns that can help us better grasp how the
popularity of a specific issue competes in the SQA.
Figure 2: Number of questions per year of topic concerns.
First, we found that some topics were not queried
in the first year, 2010. The topics that appeared for the
first time in 2011 are pitching an idea, investment
capital, scaling business, small business, important
resources, finding good sources, startup location,
well-prepared business, business model, digital
business, emerging markets, partnership, and idea
generation. However, some of the concerns receive
more attention from SEs to discuss in the SQA.
Figure 3: Dynamic changes in popularity over time of
positive concerns relative to other concerns.
Next, we present the dynamic changes in
popularity of the concerns over time (for Figure 3 to
6, on the x-axis, the years, and the y-axis, the ranking
of the concerns relative to each other based on the
number of questions asked in that year, the higher the
ranking, the more questions asked relative to other
concerns in that year). Figure 3 shows the top positive
concerns based on Table 1 (pitching, finding good
resources, idea generation, marketing advice,
Examining Competing Entrepreneurial Concerns in a Social Question and Answer (SQA) Platform
149
disruptive innovation, and keys to success), and
Figure 4 shows the top negative concerns and those
that have changed significantly over time (Figures 5
and 6).
Figure 4: Dynamic changes in popularity over time of
negative concerns relative to other concerns.
As shown in Figure 3, the topic of concern about
disruptive innovation has received more attention
from 2015 to 2018 than other concerns that show
more positive sentiment. Meanwhile, the topics of
marketing advice and finding good sources fall from
the top in 2017.
Figure 5: Concerns that show increasing trends in
popularity compared to other concerns.
Moreover, we can observe in Figure 3 that there
are significant shifts in all concerns in 2014. On the
other hand, in Figure 4, in 2015, there were no
concerns about the "founder figure," Three years
later, the concern was the highest in the group of
negative concerns. Thus, overall, the negative
concerns group did not show a consistent significant
trend over time.
An interesting pattern that shows improvement
over time is presented in Figure 5 regarding the
concerns of the inspiring company, investment
capital, scaling the business, small business owner,
and startup location. Those concerns consistently
show positive tendencies and can be a signal for
providing content learning materials in that regard.
Also, four topics of concern show negative
popularity among the graphics: lean startup, user-
centered app, product launch, and technology trends.
Many reasons might push the concern popularity to
be reduced, for instance, the availability of learning
sources outside the SQA.
Figure 6: Concerns that show decreasing trends in
popularity compared to other concerns.
The following section will discuss the study's
implications, including limitations and
recommendations for further study.
5 DISCUSSION
Based on the identified concerns from the SQA,
further research can use the proposed designation and
validate it empirically with SEs based on different
types and sizes of businesses, SEs characteristics
(Chandra et al., 2016), or the current entrepreneurial
journey process. To fit the context of the SEs study,
the list of competing concerns may be prioritized,
modified, combined, or, if necessary, expanded to
include new concerns.
In comparison to earlier research (Giardino et al.,
2015; Wang et al., 2016), this study discovered two
new dimensions: role model and international startup
development. Pitching and product launch were also
KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems
150
discussed in the SQA, which shows the importance of
startups in the digital age in terms of customer
acquisition (Giardino et al., 2015). Technology trends
and digital business also shows that the challenges
were also prevalent for SEs in the digital age, which
was previously discovered only from the social media
data of SE status on Twitter (Saura et al., 2019), but
from the direct interview and observation study
(Giardino et al., 2015; Wang et al., 2016).
Furthermore, the study demonstrates how topic
modeling and sentiment analysis may be used to
uncover SE concerns based on questions posted on an
SQA. We also confirmed SEs' usage of SQA in an
open virtual community forum to find preferences
and responses to their bewilderment, challenges, and
queries as concerns, which had previously been
generated by qualitative methods such as surveys,
literature studies, and personal interviews with SEs
(Giardino et al., 2015; Ratinho et al., 2020; Shneor &
Flåten, 2015; Wang et al., 2016).
Nonetheless, this study has a few limitations.
First, we only studied one SQA that uses English.
Further research can be conducted with other SQA
platforms and combine multilingual data collections.
Current research on language translation enables
topic modeling to mature in English by translating the
other language into English before applying the
English-based topic modeling. Also, we only looked
at the questions and excluded the answers. By
supplementing the dataset with responses, additional
information can be gathered that will aid in
determining the size of the issue at hand.
Furthermore, topics may overlap in meaning or
have almost identical meanings; since we only use the
log-likelihood value to select the number of topics,
other criteria (e.g., coherence value) can also be used.
In further studies, the related questions can be
included in the analysis to form the label and validate
the label with experts and SE directly. Besides, while
the study was limited to 2018, based on the identified
topics, the LDA model presented in Table 1 with
words and the associated coefficient can predict a
topic in new documents or questions. Future research
may indicate dynamic shifts in the popularity of
concern about a particular event, such as before and
after the Covid 19 pandemic.
6 CONCLUSION
Our research explores the SEs concern pattern and
shows that SQA can be reliable knowledge
management and entrepreneurship study tool. We
have identified thirty competing concerns that are
coupled with SEs' emotional preferences. In the
future, it is possible to conduct more empirical
explorations based on the thirty issues raised.
ACKNOWLEDGEMENTS
The first author has received a grant from the KPM
Committee of the Hochschule Ruhr West to publish
the paper with the ID: KPM321024. Additionally,
thanks to the Ministry of Culture and Science of the
State of North Rhine-Westphalia for the financial
support of the Institute of Positive Computing at the
Hochschule Ruhr West.
REFERENCES
Blanchflower, D. G., & Oswald, A. J. (1998). What Makes
An Entrepreneur? Journal Of Labor EconomicS, 16(1),
26–60.
Blei, D. M., Ng, A. Y., & Jordan, M. I. (2003). Latent
dirichlet allocation. The Journal of Machine Learning
Research, 3, 993–1022.
Chandra, Y., Jiang, L. C., & Wang, C.-J. (2016). Mining
social entrepreneurship strategies using topic modeling.
PloS One, 11(3), e0151342.
Chang, J., Boyd-Graber, J., Wang, C., Gerrish, S., & Blei,
D. M. (2009). Reading tea leaves: How humans
interpret topic models. In Neural information
processing systems (pp. 288–296). Citeseer.
Chen, Y., Yuan, J., You, Q., & Luo, J. (2018). Twitter
sentiment analysis via bi-sense emoji embedding and
attention-based lstm. In Proceedings of the 26th acm
international conference on multimedia.
Choi, E., Kitzie, V., & Shah, C. (2014). Investigating
motivations and expectations of asking a question in
social q&a. First Monday.
Gao, W., Li, S., Lee, S. Y. M., Zhou, G., & Huang, C.-R.
(2013). Joint learning on sentiment and emotion
classification. In Proceedings of the 22nd acm
international conference on information & knowledge
management.
Giardino, C., Bajwa, S. S., Wang, X., & Abrahamsson, P.
(2015). Key challenges in early-stage software startups.
In International conference on agile software
development (pp. 52–63). Springer.
Gifford, S. (1992). Allocation of entrepreneurial attention.
Journal of Economic Behavior & Organization, 19(3),
265–284.
Harb, Y., & Shang, Y. (2021). Linking information
technology and entrepreneurship: A literature review.
In Proceedings of the 54th hawaii international
conference on system sciences.
Huber, M., Mulazzani, M., Leithner, M., Schrittwieser, S.,
Wondracek, G., & Weippl, E. (2011). Social snapshots:
Digital forensics for online social networks. In
Examining Competing Entrepreneurial Concerns in a Social Question and Answer (SQA) Platform
151
Proceedings of the 27th annual computer security
applications conference.
Hutto, C., & Gilbert, E. (Eds.) (2014). Vader: A
parsimonious rule-based model for sentiment analysis
of social media text. : Vol. 8.
Jiang, H., Qiang, M., Zhang, D., Wen, Q., Xia, B., & An, N.
(2018). Climate change communication in an online
q&a community: A case study of quora. Sustainability,
10(5), 1509.
Johnson, B. R. (1990). Toward a multidimensional model
of entrepreneurship: The case of achievement
motivation and the entrepreneur. Entrepreneurship
Theory and Practice, 14(3), 39–54.
Kumar, A., Praveen, S., Goel, N., & Sanwal, K. (2018).
Opinion extraction from quora using user-biased
sentiment analysis. In Information systems design and
intelligent applications (pp. 219–228). Springer.
Nurhas, I., Geisler, S., Ojala, A., & Pawlowski, J. (Eds.)
(2020). Towards a Wellbeing-driven System Design for
Intergenerational Collaborative Innovation: A
Literature Review.
O’Shea, D., Buckley, F., & Halbesleben, J. (2017). Self-
regulation in entrepreneurs: Integrating action,
cognition, motivation, and emotions. Organizational
Psychology Review, 7(3), 250–278.
Onan, A., Korukoglu, S., & Bulut, H. (2016). Lda-based
topic modelling in text sentiment classification: An
empirical analysis. Int. J. Comput. Linguistics Appl.,
7(1), 101–119.
Puhakka, V., & Ojala, A. (2021). A review of international
entrepreneurship as part of broader opportunity
research: Topic modelling approach. Empirical
International Entrepreneurship, 89–113.
Ratinho, T., Amezcua, A., Honig, B., & Zeng, Z. (2020).
Supporting entrepreneurs: A systematic review of
literature and an agenda for research. Technological
Forecasting and Social Change, 154, 119956.
Rehurek, R., & Sojka, P. (Eds.) (2010). Software
framework for topic modelling with large corpora.
Citeseer.
Saura, J. R., Palos-Sanchez, P., & Grilo, A. (2019).
Detecting indicators for startup business success:
Sentiment analysis using text data mining.
Sustainability, 11(3), 917.
Schöning, M. (2013). Social entrepreneurs as main drivers
of social innovation. In Social innovation (pp. 111–
118). Springer.
Shneor, R., & Flåten, B.-T. (2015). Opportunities for
entrepreneurial development and growth through online
communities, collaboration, and value creating and co-
creating activities. In Entrepreneurial challenges in the
21st century (pp. 178–199). Springer.
van Stel, A., Carree, M., & Thurik, R. (2005). The effect of
entrepreneurial activity on national economic growth.
Small Business Economics, 24(3), 311–321.
Wang, G., Gill, K., Mohanlal, M., Zheng, H., & Zhao, B. Y.
(2013). Wisdom in the social crowd: An analysis of
quora. In Proceedings of the 22nd international
conference on world wide web.
Wang, X., Edison, H., Bajwa, S. S., Giardino, C., &
Abrahamsson, P. (2016). Key challenges in software
startups across life cycle stages. In International
conference on agile software development (pp. 169–
182). Springer, Cham.
KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems
152
