A Glimpse into the State and Future of (Big) Data Analytics in Austria

Results from an Online Survey

Ralf Bierig

, Allan Hanbury

, Martina Haas

, Florina Piroi

Helmut Berger

, Mihai Lupu

and Michael Dittenbach

Institute of Software Technology and Interactive Systems, Vienna University of Technology,

Favoritenstr. 9-11/188-1, A-1040 Vienna, Austria

max.recall information systems GmbH, K

unstlergasse 11/1, A-1150 Vienna, Austria

Keywords:

Data Analysis, Data Analytics, Big Data, Questionnaire, Survey, Austria.

Abstract:

We present results from questionnaire data that were collected from leading data analytics experts in Austria.

The online survey addresses very current and pressing questions in the area of (big) data analysis. Our ﬁndings

provide valuable insights about what top Austrian data scientists think about data analytics, what they consider

as important application areas that can beneﬁt from big data and data processing, the challenges of the future

and how soon these challenges will become important, and the potential research topics of tomorrow. We

visualize results, summarize our ﬁndings and suggest a possible roadmap for future decision making.

1 INTRODUCTION

The time has come for continuous and large-scale data

analytics as our digital lifes now generate the impres-

sive amount of 200 exabytes of data each year. This

is equivalent to the volume of 20 million Libraries of

Congress (Dandawate, 2013). In 2012 each Internet

minute has witnessed 100,000 tweets, 277,000 Face-

book logins, 204 million email exchanges, and more

than 2 million search queries ﬁred to satisfy our in-

creasing hunger for information (Temple, 2013).

This trend is accelerated technologically by de-

vices that primarily generate digital data without the

need for any intermediary step to ﬁrst digitize ana-

log data (e.g. digital cameras vs. ﬁlm photography

combined with scanning). Additional information is

often automatically attached to the content (e.g. the

exchangeable image ﬁle format ’Extif’) that gener-

ates contextual metadata on a very ﬁne-grained level.

This means, when exchanging pictures, one also ex-

changes his or her favorite travel destination, time

(zone), speciﬁc camera conﬁguration and the light

conditions of the place with more to come as devices

evolve. Such sensors lead to a ﬂood of machine-

generated information that create a much higher spa-

tial and temporal resolution than possible before. This

’Internet of Things’ turns previously data-silent de-

vices into autonomous hubs that collect, emit and pro-

cess data at a scale that make it necessary to have

automated information processing and analysis (Dan-

dawate, 2013) to extract more value from data than

possible with manual procedures. Today’s enterprises

are also increasing their data volumes. For exam-

ple, energy providers now receive energy consump-

tion readings from Smart Meters on a quarter-hour

basis instead of once or twice per year. In hospitals it

is becoming common to store multidimensional med-

ical imaging instead of ﬂat high-resolution images.

Surveillance cameras and satellites are increasing in

numbers and generate output with increasingly higher

resolution and quality. Therefore, the focus today is

on discovery, integration, consolidation, exploitation

and analysis of this overwhelming data (Dandawate,

2013). Paramount is the question of how all this (big)

data should be analyzed and put to work. Collecting

data is not an end but a means for doing something

sensible and beneﬁcial for the data owner, the busi-

ness and the society at large. Technologies to harvest,

store and process data efﬁciently have transformed

our society and interesting questions and challenges

have emerged of how society should handle these op-

portunities. While people are generally comfortable

with storing large quantities of personal data remotely

in a cloud there is also rising concern about data own-

ership, privacy and the dangers of data being inter-

cepted and potentially misused (Boyd and Crawford,

2012).

In this paper, we present results of a study (Berger

178

Bierig R., Hanbury A., Haas M., Piroi F., Berger H., Lupu M. and Dittenbach M..

A Glimpse into the State and Future of (Big) Data Analytics in Austria - Results from an Online Survey.

DOI: 10.5220/0004999101780188

In Proceedings of 3rd International Conference on Data Management Technologies and Applications (DATA-2014), pages 178-188

ISBN: 978-989-758-035-2

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

et al., 2014) that was conducted between June 2013

and January 2014 on the topic of (big) data analy-

sis in Austria. Speciﬁcally, we present and highlight

results obtained from an online survey that involved

leading data scientists from Austrian companies and

the public sector. The questionnaire was targeted to

identify the status quo of (big) data analytics in Aus-

tria and the future challenges and developments in this

area. We surveyed opinion from 56 experts and asked

them about their understanding of data analytics and

their projections on future developments and future

research.

The paper ﬁrst discusses related work in the next

section before describing the method that was used for

creating the questionnaire and for collecting and ana-

lyzing the feedback in section 3. Results are presented

and discussed in section 4. In section 5 we conclude

and summarize our ﬁndings and suggest actions that

are based on our ﬁndings.

2 RELATED WORK

Many countries and regions are currently developing

strategies for dealing with Big Data. Prominent exam-

ples are the consultation process to create a Public-

Private Partnership in Big Data currently underway

in Europe

, work by the National Institute of Stan-

dards and Technology (NIST) Big Data Public Work-

ing Group

as well as other groups (Agrawal et al.,

2012) in the USA, and the creation of the Smart Data

Innovation Lab

in Germany.

The recent and representative McKinsey re-

port (Manyika et al., 2013) estimates the potential

global economic value of Big Data analytics between

$3.2 trillion to $5.4 trillion every year. This value

arises by intersecting open data with commercial data

and thus providing more insights for customised prod-

ucts and services and enabling better decision mak-

ing. The report identiﬁed the seven areas of edu-

cation, transportation, consumer products, electricity,

oil and gas, healthcare and consumer ﬁnance. We ex-

panded this selection by speciﬁcally focusing on the

Austrian market and its conditions before prompting

participants with a comprehensive selection of appli-

cation areas as described later in section 4.3.

Many other surveys have been conducted on the

topic of big data and big data analytics by consult-

ing companies, but these surveys usually concentrate

http://europa.eu/rapid/press-release SPEECH-13-

893 en.htm

http://bigdatawg.nist.gov

http://www.sdil.de/de/

on large enterprises

. A summary of the 2013 sur-

veys is available (Press, 2013). A survey among peo-

ple deﬁning themselves to be Data Scientists has also

been done to better deﬁne the role of Data Scientists

(Harris et al., 2013). In this paper, a survey that takes

the views of mostly academic scientists working in

multiple areas related to data analytics is presented,

and hence provides an unusual “academic” view of

the area.

3 METHOD

Surveys are powerful tools when collecting opinion

from the masses. Our main objective was to further

specify our understanding of data analytics in Aus-

tria and to identify future challenges in this emerging

ﬁeld.

We followed the strategy of active sampling. The

identiﬁcation of Austrian stakeholders in data ana-

lytics formed the starting point: We ﬁrst scanned

and reviewed Austrian industry and research institu-

tions based on their activities and research areas. We

then identiﬁed key people from these institutions and

asked about their opinions, attitudes, feedback and

participation during a roadmapping process.

Our ﬁnal contact list comprised 258 experts,

all of them senior and visible data scientists, that

we contacted twice and invited them to complete

our questionnaire. This means our contact list has

consensus-quality and represents the current situation

and strength of senior data scientists in Austria. The

survey was online between the beginning of Septem-

ber 2013 until the middle of October 2013. A total of

105 people followed the link to the survey resulting

in a general response rate of 39%. However, several

of them turned down the questionnaire or cancelled

their efforts after only one or two questions. We took

a strict measure and removed those incomplete cases

from the list of responses to increase the quality of the

data. This reduced the original 105 responses (39%)

further down to 56 responses (21.7%).

The general advantages of online surveys, such

as truthfulness, increased statistical variation and im-

proved possibilities for data analysis (e.g. (Batinic,

2003; D

oring, 2003)), unfortunately suffer from the

problems of limited control, a higher demand on par-

ticipants in terms of time and patience and the poten-

Some examples: http://www-935.ibm.com/services/

us/gbs/thoughtleadership/ibv-big-data-at-work.html,

http://www.sas.com/resources/whitepaper/wp 58466.pdf,

the Computing Research Association (CRA)

http://www.cra.org/ccc/ﬁles/docs/init/bigdatawhitepaper.pdf

and SAS http://www.sas.com/resources/whitepaper/

wp 55536.pdf

AGlimpseintotheStateandFutureof(Big)DataAnalyticsinAustria-ResultsfromanOnlineSurvey

179

tial that people may be engaged in other, distracting

activities that alter the results and increase the dropout

rate (e.g. (Birnbaum, 2004)). While our response rate

of nearly 40% is normal for online surveys (Batinic,

2003), the high dropout rate in our speciﬁc case can

be attributed to the complex nature of the subject.

The data of the survey was collected anonymously

with LimeSurvey

and was analyzed with R, a statis-

tical software package

4 SURVEY RESULTS AND

DISCUSSION

This section highlights the results we obtained from

the data and focuses on four areas. First, the demo-

graphic information about the participants (e.g. their

age, area of work, and their work experience) that

helps us to get a better understanding of the charac-

teristics of a typical data scientist in Austria. Sec-

ond, we look at the application areas of data analytics

and how participants projected the future relevance of

these areas. Third, we investigated the future chal-

lenges of data analytics. Forth, we analysed free text

submissions from questions about research priorities

and the need for immediate funding to get a better un-

derstanding of possible future directions, interests and

desires. We omitted replies for the questions 9, 13

and 16 that inquired about other application areas and

additional general comments (see appendix 5. These

questions only had a very limited text response that

would not be meaningful to analyze statistically.

4.1 Participants

The data presented in this paper is based on the opin-

ions of 56 people who completed the questionnaire

— four female (7.1%) and 52 male (92.9%). This

gender distribution is similar in the original contact

list — 26 female (10.1%) and 232 male (89.9%) —

and therefore represents the current gender situation

in the data science profession in Austria. Partici-

pants were mostly Austrians (96%) and the majority

of them were working in the research and academic

sector. About a ﬁfth (21.4%) of all responses came

from the industry. The larger part worked for aca-

demic (55.4%) or non-industry (33.9%) research or-

ganisations

. The majority of participants (80.3%)

had an extended experience of nine or more years.

http://www.limesurvey.org/de/

http://www.r-project.org/

Multiple selections were possible which means that

these numbers do not add up to 100%.

This deﬁnes our sample as a group of mostly aca-

demic, male, and Austrian data scientists.

4.2 What is Data Analytics?

We asked participants to describe the term ’Data An-

alytics’ in their own words as an open question to get

an idea about the dimensions of the concept and the

individual views on the subject. Figure 1 depicts a

summary word cloud from the collected free-text re-

sponses for all those terms that repeatedly appeared

in the response

. It further depicts a small set of

representative extracts from the comments and deﬁ-

nitions that participants submitted. Overall, the com-

ments where very much focused on the issue of large

data volumes, the process of knowledge extraction

with speciﬁc methods and algorithms and the aggre-

gation and combination of data in order to get new

insights. Often it was related to machine learning and

data mining but as a wider and more integrative ap-

proach. Only very few respondents labeled Data An-

alytics to be simply a modern and fashionable word

for data mining or pattern recognition.

4.3 Important Application Areas

Based on the literature review that preceded this sur-

vey, we identiﬁed the main application areas of data

analytics in Austria as healthcare, commerce, man-

ufacturing and logistics, transportation, energy and

utilities, the public sector and the government, educa-

tion, tourism, telecommunication, e-science, law en-

forcement, and ﬁnance and insurance. Figure 2 shows

the relative importance of these areas as attributed by

participants. Selections were made in binary form

with multiple selections possible. The ﬁgure shows

that the area of healthcare is perceived as a strong

sector for data analytics (66.1%) followed by energy

(53.6%), manufacturing and e-science (both 50.0%).

As a sector that is perceived to beneﬁt only little from

(big) data analytics are tourism and commerce (both

23.2%). This is despite the fact that these areas are

large in Austria based on demographic data as pro-

vided by the Austrian department of Statistics

We only included terms that appeared a least three

times and we ﬁltered with an english and a topical stop word

list (e.g. terms like ’and’ or ’etc’ and terms like ’data’ or

’analytics’).

In 2010, 19.3% of the employed worked in com-

merce and 9.1% in the gastronomical and leisure sec-

tor (source: ’Ergebnisse im Ueberblick: Statistik zur

Unternehmensdemograﬁe 2004 bis 2010’, available at

http://www.statistik.at/web de/statistiken/unternehmen

arbeitsstaetten/arbeitgeberunternehmensdemograﬁe/

index.html, extracted July 15, 2014.

DATA2014-3rdInternationalConferenceonDataManagementTechnologiesandApplications

180

amounts

knowledge

unstructured

extracting

insights

mining

processing

structured

learning

methods

extraction

machine

purpose

relevant

statistical

systems

techniques

tools

analyse

base

collected

collections

extract

generating

insight

meaning

modelling

models

process

science

semantic

specific

structure

various

visualization

"Aggregating, processing, modeling and

visualizing large amounts of unstructured

and/or structured data for

knowledge extraction and decision making."

"Exploring relationships in

data on the basis of

mathematical-statistical

models."

"Theory, methods,

representations,

and implementation

of processes

for data-driven

action support."

"Extracting meaningful information

out of large amounts of

(unstructured) data"

"The technologies and

methods used to

analyse large amounts

of data

in a way which might

not related to the

primary or intended

use of these data.."

Figure 1: What participants understood as data analytics.

Healthcare

Energy

Manufacturing

E−Science

Telecommunication

Transportation

Education

Finance

Government

Commerce

Tourism

Law Enforcement

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

Figure 2: Important application areas for data analytics.

AGlimpseintotheStateandFutureof(Big)DataAnalyticsinAustria-ResultsfromanOnlineSurvey

181

Healthcare Commerce Manufacturing Transportation

Energy Government Education Tourism

Telecommunication E−Science Law Enforcement Finance

short term

medium term

long term

not important

do not know

short term

medium term

long term

not important

do not know

short term

medium term

long term

not important

do not know

10%

20%

30%

40%

10%

20%

30%

40%

10%

20%

30%

40%

10%

20%

30%

40%

Figure 3: Application areas where data analytics will become important in the short, middle and long term.

We additionally asked participants to provide fu-

ture projections for these application areas w.r.t. how

they think these areas would become important for

data analytics in the future (see ﬁgure 3). Here, par-

ticipants rated the application areas based on their rel-

evance for the short, middle and long term future. The

diagram also visualizes the amount of uncertainty in

these projections as participants could select if they

were unsure or even declare an application area as

unimportant. The ﬁgure shows that application areas

that are perceived as strong candidates (e.g. health-

care, energy and telecommunication) are all marked

as relevant for the short term future with decreasing

ratings on the longer timeline. Less strongly per-

ceived application areas, such as law enforcement and

tourism have results that are less clearly expressed

with a stronger emphasis on a longer time frame. The

amount of uncertainty about these areas is also much

higher. Law enforcement is perceived as both less im-

portant and not beneﬁting from data analytics. This

is conceivable as law enforcement may not be per-

ceived as an independent sector, as this is the case in

the United States (Norton, 2013) where data analyt-

ics already assists the crime prediction process with

data mining, e.g. with the use of clustering, clas-

siﬁcation, deviation detection, social network analy-

sis, entity extraction, association rule mining, string

comparison and sequential pattern mining. It comes a

bit as a surprise that tourism in Austria was both per-

ceived as rather unimportant and also as an area that

would only beneﬁt from data analytics in the mid- and

long-term future. The large proportion of uncertainty

DATA2014-3rdInternationalConferenceonDataManagementTechnologiesandApplications

182

Privacy/Security Algorithm Issues Qualified Personnel

Data Preservation Evaluation Data Ownership

short term

medium term

long term

not important

do not know

short term

medium term

long term

not important

do not know

10%

20%

30%

40%

50%

10%

20%

30%

40%

50%

10%

20%

30%

40%

50%

Figure 4: Future challenges in data analytics for the short, middle and long term.

shows that experts seem to be rather unsure about the

future of these two sectors.

4.4 Current and Future Challenges

Based on the literature review, we identiﬁed the main

challenges of data analytics in the areas of privacy

and security, algorithmic and scalability, getting qual-

iﬁed personnel, the preservation and curation process,

the evaluation and benchmarking, and data ownership

and open data. We now asked participants to catego-

rize these challenges into three groups: Short term if

they see it as an issue of the very near future, medium

term if there is still time and long term if this might

become an issue some time in the far future. Our in-

tent was to obtain a priority that can help us to iden-

tify possible actions and recommendations for deci-

sion making. Figure 4 depicts all responses for all

categories and also includes the amount of uncertainty

(do not know) and how unimportant people thought it

to be (not important).

All challenges share that they are perceived as all

being relevant in the short and middle-term future,

with high certainty throughout. Upon closer inves-

tigation, the response can then be further divided into

two groups.

The ﬁrst group of challenges consists of privacy

and security and the issue of qualiﬁed personnel.

These issues are perceived as considerably more im-

portant and pressing in the short term future than the

middle and the long term future. It was especially

striking how important privacy and security was em-

AGlimpseintotheStateandFutureof(Big)DataAnalyticsinAustria-ResultsfromanOnlineSurvey

183

Table 1: Comparison of Big Data Challenges as identiﬁed in three different studies.

CRA Study SAS Study Our Study

Lack of Skills / Experience - Personnel

Accessing data / Sharing Human Collaboration Data Ownership / Data Preservation

Effective Use - -

Analysis and Understanding Data Heterogenity Algorithm Issues

Scalibility Scalibility/Timeliness Evaluation

- Privacy Privacy and Security

phasised throughout the entire study — including this

online questionnaire. This strong impact might be

partially attributed to the very recent NSA scandal.

However, it might also be that our target group quite

naturally possesses a heightened sensitivity to the po-

tential dangers of (big) data analytics and the often

unprotected ﬂow of personal data on the open web.

Qualiﬁed personnel is a problem of the near future

and has been discussed in the literature throughout

many studies. This is well conﬁrmed in our own ﬁnd-

ings and an important issue to address in future deci-

sion making.

The second group of challenges covers algorith-

mic and scalability issues, data preservation and cu-

ration, evaluation, and data ownership and open data.

All of these were attributed more frequently to be is-

sues of the future. Ironically, data preservation and

curation has been attributed with being more relevant

in the long-term future than the mid- and short-term

with the highest amount of uncertainty in the entire

response. This should ideally be the opposite. We

would have also expected that data ownership and

open data issues would be categorized very similar

to the privacy and security response and that the algo-

rithmic issues are more relevant on a short term scale

as data is mounting very fast. The responses never-

theless demonstrate the feeling that the privacy and

security and qualiﬁed personnel challenges need to be

solved before progress can be made in the ﬁeld.

We additionally compared our list of challenges

with those that were identiﬁed in two related, recent

studies: One study hosted by the Computing Research

Association (CRA) that focused on Challenges and

Opportunities of Big Data in general

and one study

on Big Data visualization by SAS

. In Table 1, we

refer to them as ’CRA Study’, ’SAS Study’ and ’Our

Study’ and compare 6 challenge categories that were

identiﬁed across these studies. The challenges are

presented in no particular order, however, the reader

can compare challenge categories horizontally in the

http://www.cra.org/ccc/ﬁles/docs/init/

bigdatawhitepaper.pdf

http://www.sas.com/resources

/whitepaper/wp 55536.pdf

table. A dash (-) means that a particular challenge

was not identiﬁed by a study. We related the cate-

gories with each other to give the reader an overview

about the similarities and differences from three per-

spectives. Naturally, the categories did not always

represent a perfect match. For example, the challenge

of data access and data sharing was addressed as the

need for human collaboration in the SAS study and

our own study identiﬁed the challenge of data owner-

ship and the challenge of preserving data in this cat-

egory. However, the issue of a lack of personnel was

also identiﬁed as a lack of skills and experience in the

CRA study. Whereas privacy was clearly addressed in

the SAS study and more comprehensively combined

with security in our study, the CRA study did not con-

sider it a challenge at all. Overall, this comparison

shows that there is considerable agreement between

studies with respect to future challenges. It would

be interesting to further extend this comparison to a

much wider range of studies in future work.

4.5 Future Research Topics

We prompted participants with two questions about

future research in (big) data analytics. The ﬁrst ques-

tion asked them to enter free text on topics of their

preferred future research which they had to prioritise

by three levels (top priority, 2nd priority, 3rd prior-

ity). The second question can be seen as a reﬁne-

ment of the top priority level of the previous question

and asked them to describe which research topic they

would like to see publicly funded with 10 Million .

Again, this was submitted as free text allowing par-

ticipants to contribute their ideas in a completely free

and unrestricted form. Figure 5 shows the term fre-

quencies of those texts for all three priorities and also

the text for the 10 Million research topic. This allows

for easy comparisons. A sum across all four columns

of frequencies provides an overview about the entire

topic space.

The most frequent themes are privacy/security

(mentioned 22 times) and healthcare (mentioned 17

times) which coincides with the ﬁndings from the pre-

vious questions. The importance of privacy and secu-

rity was found to be the most pressing future chal-

DATA2014-3rdInternationalConferenceonDataManagementTechnologiesandApplications

184

algorithms

healthcare

energy

learning

machine

management

modelling

privacy and security

processing

scalable

semantic

simulation

support

systems

visual

applications

education

environment

pattern

preservation

tools

distributed

infrastructure

Top Priority

2nd Priority

3rd Priority

10 Mio Euro

Sum

Priority

Term

Figure 5: Priorities for future research as expressed by participants.

lenge (see Figure 4). Likewise, healthcare was also

perceived as the most promising application domain

(see Figure 2) with the most pressing time line that

strongly leans toward the short-term future (see Fig-

ure 3). The third most frequent keyword were seman-

tic issues (mentioned 16 times) that were more exten-

sively investigated in a number of workshops and an

expert interview that is documented in more detail in

(Berger et al., 2014).

5 CONCLUSIONS AND

TECHNOLOGY ROADMAP

This paper presented results from a study on (big) data

analysis in Austria. We summarized the opinons of

Austrian data scientists sampled from both industry

and the academia and some of the most pressing and

current issues in the ﬁeld.

We found that data analytics is understood as deal-

ing with large data volumes, where knowledge is ex-

tracted and aggregated to lead to new insights. It

was interesting to see that it was often related to data

mining but viewed more widely and more highly in-

AGlimpseintotheStateandFutureof(Big)DataAnalyticsinAustria-ResultsfromanOnlineSurvey

185

tegrated. Healthcare was seen as the most impor-

tant application area (66.1%), followed by energy

(53.6%), manufacturing and logistics (50.0%) and e-

science (50.0%) with big potential in the short term

future. Other areas were judged less important, such

as tourism (23.2%) and law enforcement (17.9%),

with high uncertainty. We found that our literature-

informed list of challenges were conﬁrmed by our

respondents, however only privacy/security and the

challenge to get qualiﬁed personnel was strongly

attributed to the very near future. Algorithm is-

sues, data preservation, evaluation and data owner-

ship were seen as challenges that become more rel-

evant only in the longer run. Research priorities

and funding requests where strongly targeted to pri-

vacy/security (mentioned 22 times), healthcare (men-

tioned 17 times) and semantic issues (mentioned 16

times). This result conforms largely to the ﬁndings in

the other parts of the study.

Based on the results of the survey presented in this

paper, along with the outcomes of three workshops

and interviews, a technology roadmap consisting of a

number of objectives was drawn up. The roadmap ac-

tions are described in much greater detail in (Berger

et al., 2014) as part of the complete report that focuses

on all parts of the study. This is outside the scope of

this paper that focuses on the details of the online sur-

vey. In summary, the identiﬁed challenges, together

with their careful evaluation, have led to three cate-

gories of actions that are manifested in this roadmap.

First, to meet the challenges of data sharing,

evaluation and data preservation, an objective in the

roadmap is to create a “Data-Services Ecosystem” in

Austria. This is related to an objective to create a le-

gal and regulatory framework that covers issues such

as privacy, security and data ownership, as such a

framework is necessary to have a functioning Ecosys-

tem. In particular, it is suggested to fund a study

project to develop the concept of such an Ecosystem,

launch measures to educate and encourage data own-

ers to make their data and problems available, and

progress to lighthouse projects to implement and re-

ﬁne the Ecosystem and its corresponding infrastruc-

ture. Furthermore, it is recommended to develop a

legal framework and create technological framework

controls to address the pressing challenges of privacy

and security in data analytics.

Second, technical objectives are to overcome chal-

lenges related to data integration and fusion and algo-

rithmic efﬁciency, as well as to create actionable in-

formation and revolutionise the way that knowledge

work is done. We suggest to fund research that fo-

cuses on future data preservation, to develop fusion

approaches for very large amounts of data, to create

methods that assure anonymity when combining data

from many sources, to enable real time processing,

and to launch algorithmic challenges. A full list of

suggestions are described in more detail in (Berger

et al., 2014).

The third and ﬁnal objective in the roadmap is to

increase the number of data scientists being trained.

We suggest a comprehensive approach to create these

human resources and competences through educa-

tional measures at all levels: from schools through

universities and universities of applied sciences to

companies. The issue of having more and highly

skilled data scientists soon is an issue that requires

immediate action to secure the future prosperity of the

Austrian (big) data analytics landscape.

ACKNOWLEDGEMENTS

This study was commissioned and funded by the

Austrian Research Promotion Agency (FFG) and the

Austrian Federal Ministry for Transport, Innovation

and Technology (BMVIT) as FFG ICT of the Fu-

ture project number 840200. We thank Andreas

Rauber for his valuable input. Information about the

project and access to all deliverables are provided at

http://www.conqueringdata.com/.

REFERENCES

Agrawal, D., Bernstein, P., Bertino, E., Davidson, S., Dayal,

U., Franklin, M., Gehrke, J., Haas, L., Halevy, A.,

Han, J., Jagadish, H. V., Labrinidis, A., Madden, S.,

Papakonstantinou, Y., Patel, J. M., Ramakrishnan, R.,

Ross, K., Shahabi, C., Suciu, D., Vaithyanathan, S.,

and Widom, J. (2012). Challenges and Opportunities

with Big Data. http://www.cra.org/ccc/ﬁles/docs/init/

bigdatawhitepaper.pdf, last visited: August 2013.

Batinic, B. (2003). Internetbasierte Befragungsverfahren.

Osterreichische Zeitschrift f

ur Soziologie, 28(4):6–18.

Berger, H., Dittenbach, M., Haas, M., Bierig, R., Hanbury,

A., Lupu, M., and Piroi, F. (2014). Conquering Data

in Austria. bmvit (Bundesministerium f

ur Verkehr, In-

novation and Technology, Vienna, Austria.

Birnbaum, M. H. (2004). Human research and data col-

lection via the internet. Annual review of Psychology,

55:803–832.

Boyd, D. and Crawford, K. (2012). Critical Questions for

Big Data. Information, Communication & Society,

15(5):662–679.

Dandawate, Y., editor (2013). Big Data: Challenges and

Opportunities, volume 11 of Infosys Labs Brieﬁngs.

Infosys Labs. http://www.infosys.com/infosys-

labs/publications/Documents/bigdata-challenges-

opportunities.pdf, last visited: August 2013.

DATA2014-3rdInternationalConferenceonDataManagementTechnologiesandApplications

186

oring, N. (2003). Sozialpsychologie des Internet.

Die Bedeutung des Internet f

ur Kommunikation-

sprozesse, Identit

aten, soziale Beziehungen und Grup-

pen. Hogrefe, G

ottingen, 2nd edition.

Harris, H., Murphy, S., and Vaisman, M. (2013). Analyzing

the Analyzers: An Introspective Survey of Data Scien-

tists and Their Work. O’Reilly.

Manyika, J., Chui, M., Groves, P., Farrell, D., Kuiken,

S. V., and Doshi, E. A. (2013). Open data: Unlocking

innovation and performance with liquid information.

McKinsey Global Institute.

Norton, A. (2013). Predictive Policing - The Future of Law

Enforcement in the Trinidad and Tobago Police Ser-

vice. Int. J. of Computer Applications, 62:32–36.

Press, G. (2013). The state of big

data: What the surveys say.

http://www.forbes.com/sites/gilpress/2013/11/30/the-

state-of-big-data-what-the-surveys-say/, last visited:

March 2014.

Temple, K. (2013). What Happens in an Internet

Minute? http://scoop.intel.com/what-happens-in-an-

internet-minute/, last visited: August 2013.

APPENDIX: QUESTIONS

The questions of the online survey are categorized in

background information, research and development

focus, challenges, and public funding.

Background information

Question 1: What is your current activity environ-

ment? (Provide comments if you wish to):

• Academia (University)

• Non-University Research

• Industry

• Public Ofﬁce

Question 2: How many years of experience do you

have in your activity area?

• 1-3

• 4-8

• 9 or more

Question 3: Would you consider yourself a... (Pro-

vide comments if you wish to)

• Researcher

• Service Provider

• Policy Maker

• User of Data Analytics technology

• Other:

Question 4: Gender

• Male

• Female

Question 5: In which country do you work?

• List of countries

Data Analytics Deﬁnition

Question 6: What is your understanding of Data

Analytics?

• Free-test answer

Research and Development Focus

Question 7: Which of the following sub-ﬁelds do

you focus on? (Provide speciﬁc details if you wish

to)

• Search and Analysis

• Semantic Processing

• Cognitive Systems

• Visualisation and Presentation

• Other:

Question 8: Which of the following Application

Domains do you ﬁnd important today? (i.e. Applica-

tion Domains you might already be working on. Pro-

vide speciﬁc details if you wish to)

• Healthcare

• Commerce

• Manufacturing and Logistics

• Transportation

• Energy and Utilities

• Public Sector / Government

• Education

• Tourism

• Telecommunications

• eScience (incl. Life Science)

• Law Enforcement

• Finance and Insurance

Question 9: Other Application Domains you ﬁnd

important

• Free-test answer

AGlimpseintotheStateandFutureof(Big)DataAnalyticsinAustria-ResultsfromanOnlineSurvey

187

Challenges

Question 10: Which challenges do you see in Data

Analytics?

• Free-test answer

Question 11: Following your previous answer,

please judge if the following challenges will be im-

portant in the short, medium or long term.

• Privacy and Security – short term, medium term,

long term, not important, don’t know

• Algorithm Issues (e.g. Scalability) – short term,

medium term, long term, not important, don’t

know

• Qualiﬁed Personnel – short term, medium term,

long term, not important, don’t know

• Data Preservation and Curation – short term,

medium term, long term, not important, don’t

know

• Evaluation and Benchmarking – short term,

medium term, long term, not important, don’t

know

• Data Ownership and Open Data – short term,

medium term, long term, not important, don’t

know

Question 12: Which challenges do you see in Data

Analytics?

• Healthcare – short term, medium term, long term,

not important, don’t know

• Commerce – short term, medium term, long term,

not important, don’t know

• Manufacturing and Logistics – short term,

medium term, long term, not important, don’t

know

• Transportation – short term, medium term, long

term, not important, don’t know

• Energy and Utilities – short term, medium term,

long term, not important, don’t know

• Public Sector / Government – short term, medium

term, long term, not important, don’t know

• Education – short term, medium term, long term,

not important, don’t know

• Tourism – short term, medium term, long term,

not important, don’t know

• Telecommunications – short term, medium term,

long term, not important, don’t know

• eScience (incl. Life Science) – short term,

medium term, long term, not important, don’t

know

• Law Enforcement – short term, medium term,

long term, not important, don’t know

• Finance and Insurance – short term, medium term,

long term, not important, don’t know

Question 13: Other Application Domains you ﬁnd

important (please indicate Short/Medium/Long Term)

• Free-test answer

Public Funding

Question 14: Which research areas or topics in the

Data Analytics ﬁeld are most important and should be

prioritized by public funding (name 3 and rank)

• Top Priority: Free-test answer

• Second Priority: Free-test answer

• Third Priority: Free-test answer

Question 15: Please complete the following news

headline: 10,000,000 Euro for...

• Free-test answer

Question 16: Other comments you might have

about data analytics

• Free-test answer

DATA2014-3rdInternationalConferenceonDataManagementTechnologiesandApplications

188