Application of Methodologies and Process Models in Big Data

Projects

Rosa Quelal

, Luis Eduardo Mendoza

1,2 b

and Mónica Villavicencio

ESPOL Polytechnic University, Escuela Superior Politécnica del Litoral, ESPOL,

Facultad de Ingeniería en Electricidad y Computación, Campus Gustavo Galindo Km. 30.5 Vía Perimetral,

P.O. Box 09-01-5863, Guayaquil, Ecuador

Processes and Systems Department, Simón Bolívar University, Valle de Sartenejas, P.O. Box 89000, Caracas, Venezuela

Keywords: Agile Methodologies, Big Data, Systematic Literature Review, Text Mining.

Abstract: The concept of Big Data is being used in different business sectors; however, it is not certain which

methodologies and process models have been used for the development of these kind of projects. This paper

presents a systematic literature review of studies reported between 2012 and 2017 related to agile and non-

agile methodologies applied in Big Data projects. For validating our review process, a text mining method

was used. The results reveal that since 2016 the number of articles that integrate the agile manifesto in Big

Data project has increased, being Scrum the agile framework most commonly applied. We also found that

44% of articles obtained from a manual systematic literature review were automatically identified by applying

text mining.

1 INTRODUCTION

Big Data projects have been used in different

economic sectors. Therefore, it is necessary to study

how Big Data projects are planned and executed in

order to reach their expectations —execution time

(Al-Jaroodi et al., 2017), return on investment, and

value for client (Chen et al., 2016). To do so, we

performed a Systematic Literature Review (SLR)

related to methodologies and process models applied

in Big Data projects.

From the SLR, we notice that there is an emerging

interest in applying software engineering process

models to Big Data initiatives (Al-Jaroodi et al.,

2017; Kumar, 2017); i.e., we observe a growth of

publications related to both concepts (software

engineering and big data). Therefore, in order to

improve our literary review process, which involves

the continuous incorporation of emerging

publications related to these concepts, we validate our

manual SLR process with a text mining method.

The rest of this article contains the following:

section 2 describes the research methodological

https://orcid.org/0000-0002-9292-8441

https://orcid.org/0000-0002-5081-7559

https://orcid.org/0000-0002-2601-2638

framework; section 3 shows the results obtained;

section 4 presents the limitations and future work, and

section 5 concludes the article.

2 METHODOLOGICAL

FRAMEWORK

To conduct this SLR, we used the guidelines

proposed by Kitchenham and Petersen (Kitchenham

and Charters, 2007; Petersen et al., 2015), including:

the formulation of research questions; the search

process; inclusion and exclusion criteria; data

extraction; data analysis and classification; and

quality evaluation.

2.1 Research Questions

The purpose of this research is to identify which

methodologies and process models have been applied

in Big Data projects. Hence, the research questions

are:

Quelal, R., Mendoza, L. and Villavicencio, M.

Application of Methodologies and Process Models in Big Data Projects.

DOI: 10.5220/0007729602770284

In Proceedings of the 21st International Conference on Enterprise Information Systems (ICEIS 2019), pages 277-284

ISBN: 978-989-758-372-8

277

1. What Agile Methodologies have been applied in

Big Data projects?

2. What Non-agile Methodologies have been applied

in Big Data projects?

3. What kind of Big Data projects are applying either

agile or non-agile methodologies?

2.2 Search Processes

To answer the first question, a manual search process

was carried out on the following databases: Science

Direct, Google Scholar, Springer and Scopus. The

process was conducted four times: November 2016,

April 2017, October 2017, and May 2018. For the

Science Direct, Google Scholar, and Springer

databases, the search was run on all article content;

and for Scopus, the search was run on titles,

keywords, and abstracts.

The query used was:

(“scrum” or “extreme programming” or

“agile Data” or “crystal” or “Kanban” or

“agile software”) and (“big data” or

“data science” or “analytic”) and (“case

study”)

To answer the second question, a similar process

was carried out. It was executed once, in July 2018 on

the following databases: Science Direct, Google

Scholar and Springer.

The query used was:

(“waterfall” or “spiral”) and (“big

data” or “data science” or “analytic”)

and (“case study”)

To answer the third question, we classified the

articles obtained in both queries according to the kind

of big data projects (i.e., paper S28 applied Text

Mining in Social Media).

2.3 Inclusion and Exclusion Criteria

Although the first “Agile Manifesto” was published

in 2001, this field of research shows fluctuations over

time, with increases in publications in 2005, 2010 and

2013 (Batra and Dahiya, 2016; Campanelli and

Parreiras, 2015). On the other hand, research related

to Big Data and its application starts increasing in

2012 (Gandomi and Haider, 2015; Wamba et al,

2015). Given these facts, we consider a range of years

from 2012 to 2017 to perform the queries. However,

traditional software engineering methodologies and

process models exist long time ago; that is why, we

perform the second search process without an initial

year and until 2017.

Additionally, the language for the search process

was restricted to “English”, and the type of

publication to “Scientific Articles”.

The inclusion criteria considered in the first

search process are:

• Agile methodologies.

• Big Data projects of any kind.

• Type of research reported: “Case Study”.

The inclusion criteria considered in the second

search process are:

• Non-agile methodologies and software process

models.

• Type of project Big Data.

• Type of research used “Case Study”.

2.4 Data Extraction

From the first search process (Agile Methodologies),

we found the following amount of articles:

• Science Direct: 170.

• Springer: 96

• Scopus: 11

• Google Scholar: 34

And from the second search process (Non-agile

Methodologies), the followings:

• Science Direct: 3

• Springer: 47

• Google Scholar: 13

In both search processes, the articles listed in

Google Scholar correspond to the ones not obtained

from the other databases.

By reviewing the list of the articles from both

searches, we found an intersection of 9 articles [S1,

S39, S57, S69, S162, S169, S198, and S247].

Therefore, the extraction process resulted in 365

articles, which are included in Appendix A.

2.5 Data Analysis and Classification

The data analysis and classification were carried out

based on the defined inclusion criteria and

classification steps, as follows:

1. Reading the abstracts.

2. Searching for each criterion within the complete

content of the articles.

3. When necessary, reading the whole article.

4. Classifying articles by criteria.

5. Classifying articles by research type.

Below, an example of each step:

(Step 1) While reading the abstract of the article

[S272], we identify a Big Data project and a

“Framework” type research report. However, the use

of agile methodologies was missing despite the

existence of the word “Scrum”.

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

278

(Step 2) When searching for the word “Scrum”

into the article, we find a strange coincidence, which

warns us for the need to perform a more in-depth

review.

(Step 3) When reading the whole article, we

realize that the word “SCRUM” stands for Spatio-

Chronological Rugby Union Model, without any

relation to agile methodologies.

(Step 4) We classify the article as Big Data

projects without Agile Methodologies, and

(Step 5) The type of research as a “Framework”.

In summary, step 1 allows us to detect articles

other than “Case Studies”. Steps 2 and 3 to identify

other types of research such as: interviews, literature

reviews, systematic mappings, case studies,

frameworks, and conceptual models. Steps 4 and 5 to

classify the articles. With this procedure, we verify all

criteria and avoid discarding articles. Appendices B

and C contain the list of articles by search criteria.

2.6 Quality Evaluation

The quality evaluation is performed in two ways: (1)

by following the guidelines proposed by Kitchenham

and Petersen (Kitchenham and Charters, 2007;

Petersen et al., 2015), defining a procedure for each

step of the guidelines, and (2) by using a text mining

method to validate the manual search processes.

The chosen text mining method was topic

classification, specifically the Latent Dirichlet

Allocation (LDA) algorithm from the Python’s

Gensim library. We chose this algorithm because it is

one of the most used in similar contexts (Chuang et

al., 2012). The process was developed in 3 stages.

Stage 1: Topic modelling process for 365 articles.

The steps followed are:

1. Collect the 365 articles obtained from the

extraction process.

2. Convert documents from .pdf to .txt format.

3. Tokenize documents; convert words into data to

be analysed.

4. Remove stopwords and punctuation marks.

5. Select 30 articles randomly to generate the corpus,

dictionary and models for three topics.

6. Classify 365 articles according to the generated

models.

Stage 2: Topic modelling process for 18 articles.

The steps followed were:

1. Collect the 18 articles obtained from the manual

review (i.e., sections 2.1 to 2.5).

2. Convert documents from .pdf to .txt format.

3. Tokenize documents; convert words into data to

be analysed.

4. Remove stopwords and punctuation marks.

5. Generate the corpus, dictionary and models for

three topics, with the 18 articles.

6. Classify 18 articles according to the generated

models.

Stage 3: Compare results from Stage 1 and 2.

3 RESULTS

By applying the methodology, we obtained 374

articles, 311 from the first search process and 63 for

the second. Since 9 articles appear in both searches,

we perform the analysis and discussion of results of

365 articles.

3.1 First Query Results

The classification of articles gives the following

results: 20% (62 articles) refer to Agile

Methodologies and 38% (117 articles) to Big Data

projects. Additionally, 66% (206 articles)

corresponds to Case Studies. From the 311 papers

obtained in the search, only 14 include the criteria:

“Case Studies”, “Agile Methodologies”, and “Big

Data”. Figure 1 shows a time line of papers per search

criterion. From 2016, the number of articles that

integrate Agile Methodologies and Big Data concepts

have increased.

Figure 1: Classification of papers by search criteria.

Table 1 summarizes the information of the 14

articles; four of them [S30, S40, S57, and S241]

report using Agile Methodologies in Big Data

projects without specifying which one.

Application of Methodologies and Process Models in Big Data Projects

279

Table 1: Papers about case studies of Agile Methodologies used in Big Data projects.

ID Year Reference Data Base Agile Methodologies Big Data Context

S30 2015

Berry, N. M., Prugh, W., Lunkins, C., Vega, J.,

Landry, R. and Garciga, L. (2015). Selecting

video analytics using cognitive ergonomics: A

case study for operational experimentation.

Procedia Manufacturing, 3, 5245-5252.

ScienceDirect Not specified

Video

Intelligence

Software

development

S146 2015

Kalenkova, A. A., van der Aalst, W. M.,

Lomazova, I. A. and Rubin, V. A. (2017). Process

mining using BPMN: relating event logs and

process models. Software and Systems Modeling,

16(4), 1019-1048.

Springer Kanban, Scrum

Process

mining

techniques

using event

logs

Academic -

Software

development

S158 2015

Komenda, M., Schwarz, D., Švancara, J., Vaitsis,

C., Zary, N. and Dušek, L. (2015). Practical use

of medical terminology in curriculum mapping.

Computers in biology and medicine, 63, 74-82.

ScienceDirect

eXtreme

Programming (XP)

Academic

Curriculum

management

Medical

S57 2016

Cope, B. and Kalantzis, M. (2016). Big Data

comes to school: Implications for learning,

assessment, and research. AERA Open, 2(2),

2332858416641907.

GoogleScholar Not specified

Educational

data mining

Academic -

scholastic

S28 2017

Baur, A. W. (2017). Harnessing the social web to

enhance insights into people’s opinions in

business, government and public administration.

Information Systems Frontiers, 19(2), 231-251.

Springer Scrum

Text Mining

in Social

Media

Automotive

industry

S40 2017

Bucksch, A., Das, A., Schneider, H., Merchant,

N. and Weitz, J. S. (2017). Overcoming the law

of the hidden in cyberinfrastructures. Trends in

plant science, 22(2), 117-123.

ScienceDirect Not specified

Analysis of

images of

plants

Scientific -

Biology

S52 2017

Chrimes, D. and Zamani, H. (2017). Using

Distributed Data over HBase in Big Data

Analytics Platform for Clinical

Services. Computational and Mathematical

Methods in Medicine, 2017.

GoogleScholar Agile Data Science

Analysis of

hospital data

about 9

billion

patients

Hospital

health

S241 2017

Ryan, P. J. and Watson, R. B. (2017). Research

Challenges for the Internet of Things: What Role

Can OR Play? Systems, 5(1), 24.

GoogleScholar Not specified

Analysis of

data from

IoT

Academic -

Scientific

Operations

Research

S245 2017

Saltz, J. and Crowston, K. (2017, January).

Comparing data science project management

methodologies via a controlled experiment.

In Proceedings of the 50th Hawaii International

Conference on System Sciences.

GoogleScholar Scrum, Kanban

Algorithms,

data mining

and machine

learning to

geographic

information

Academic -

University

S246 2017

Saltz, J. (2017). Acceptance Factors for Using a

Big Data Capability and Maturity Model.

GoogleScholar

They analyse agile

methodologies in Big

Data projects

Different

projects

Business

S247 2017

Saltz, J., Shamshurin, I. and Connors, C. (2017).

Predicting data science sociotechnical execution

challenges by categorizing data science

projects. Journal of the Association for Information

Science and Technology.

GoogleScholar Scrum

Different

types of

efforts in

data science

Business

S269 2017

Su, Y., Luarn, P., Lee, Y. S. and Yen, S. J. (2017).

Creating an invalid defect classification model using

text mining on server development. Journal of

Systems and Software, 125, 197-206.

ScienceDirect Scrum

Data mining

techniques

Software

development

S292 2017

Vidgen, R., Shaw, S. and Grant, D. B. (2017).

Management challenges in creating value from

business analytics. European Journal of

Operational Research, 261(2), 626-639.

GoogleScholar,

ScienceDirect

eXtreme

Programming (XP),

Scrum

Different

types of

efforts in

data science

Business

S300 2017

Woodside, A. G. and Sood, S. (2017). Vignettes

in the two-step arrival of the internet of things and

its reshaping of marketing management’s service-

dominant logic. Journal of Marketing

Management, 33(1-2), 98-110.

GoogleScholar Scrum

Analysis of

data from

IoT to

support

marketing

Business

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

280

3.2 Second Query Results

The classification process reveals that 60% (38

articles) refer to Non-agile Methodologies, 22% (14

articles) to Big Data projects, and 83% (52 articles) to

case studies. Only five articles include the three

criteria (Non-Agile Methodology, Big Data, and Case

Studies); however, three of them [S322, S344, and

S354] do not specify the methodology used –see

Table 2. As it can be observed, paper S247 appears in

both Tables (1 and 2). In Figure 2, we observe that

papers related to Big Data start increasing in 2012,

and papers integrating Non-agile Methodologies with

Big Data projects in 2014.

3.3 Answering Research Questions

What Agile Methodologies have been applied in Big

Data projects? Table 1 provides the answer to this

question, showing that Scrum, Kanban, XP and Agile

Data Science are the Agile Methodologies used in Big

Data projects reported from 2015 to 2017.

What Non-agile Methodologies have been applied

in Big Data projects? In our review, we identify that

CRISP-DM, SEMMA, and KDDM (Knowledge

Discovery via Data Analytics) are Non-agile

Methodologies used in big data projects.

Figure 3 summarizes the agile and non-agile

methodologies used in Big Data projects from 2012

to 2017.

What kind of Big Data projects are applying either

agile or non-agile methodologies? From the reviewed

articles, we observe a variety of Big Data projects in

which Agile and Non-agile Methodologies have been

applied, such as image intelligence, process mining,

text mining, association rule mining, geographic

information analysis, machine learning algorithms,

Internet of Thinks (IoT), etc. These projects were

developed in different contexts: academic, industrial,

scientific, business, banking, among others (see the

last column of Tables 1 and 2).

Table 2: Papers about case studies of Traditional Methodologies used in Big Data projects.

ID Year Reference Data Base Methodologies Big Data Context

S247 2017

Saltz, J., Shamshurin, I. and Connors, C. (2017).

Predicting data science sociotechnical execution

challenges by categorizing data science

projects. Journal of the Association for Information

Science and Technology.

GoogleScholar Crisp-DM

Different

types of

efforts in

data science

Business

S322 2015

D’Souza, M. J., Kashmar, R. J., Hurst, K., Fiedler, F.,

Gross, C. E., Deol, J. K. and Wilson, A. (2015).

Integrative biological chemistry program includes the

use of informatics tools, GIS And SAS software

applications. Contemporary issues in education

research (Littleton, Colo.), 8(3), 193.

GoogleScholar Not specified

Different

techniques

Applied to

GIS data

Biology

S340 2016

Li, Y., Thomas, M. A. and Osei-Bryson, K. M.

(2016). A snail shell process model for knowledge

discovery via data analytics. Decision Support

Systems, 91, 1-12.

ScienceDirect

Crisp-DM, SEMMA,

KDDM, KDDA

Different

techniques

Business

S344 2017

Mafereka, M. and Madikane, N. (2017) Data

Management is key to Banks’success.

GoogleScholar Not specified

Different

techniques

Banking

S354 2016

Saltz, J., Shamshurin, I. and Connors, C. (2016, July).

A Framework for Describing Big Data Projects. In

International Conference on Business Information

Systems (pp. 183-195). Springer, Cham.

GoogleScholar Not specified

Different

techniques

Figure 2: Classification of papers by search criteria.

Application of Methodologies and Process Models in Big Data Projects

281

Figure 3: Methodologies used in Big Data projects.

3.4 Quality Evaluation

To perform the quality evaluation, we develop the

process explained in section 2.6. Table 3 shows the

topics obtained from mining the 365 papers, with the

model generated from 30 aleatory papers. Each topic

represents a set of words that match the analyzed

documents. For example, the topic 1 is focused on

software development, topic 2 models and simulation,

and topic 3 on products, process and manufacture.

Table 3: Topic model generated from 365 papers.

Topic 1 Topic 2 Topic 3

use model data

model use product

develop figure use

system simulation manufacture

software electronic grind

information image machine

social result system

manage time process

waterfall value technology

data cell manage

Table 4 shows the topic modelling generated from

the 18 articles obtained from the whole manually

performed SLR. Topic 1 is about testing, topic 2

about business analytic and research, and topic 3

about big data and data science.

First, we evaluate the search process, and the

inclusion and exclusion criteria process. For this

evaluation, we compare the words used for the

manual queries (query 1 and query 2 from page 2 –

section 2.2) with the words generated by the topic

model (see Table 4). The matching five words appear

in light blue:

data

analytic

software

big

, and

science

Table 4: Topic model generated from manual SLR.

Topic 1 Topic 2 Topic 3

defect data data

test system project

develop research process

use use model

material analytic cid

project model team

software business use

model manage big

manuscript develop science

function inform system

Compound’s ID number, used in [48]

This fact makes us to infer that a simpler query

(i.e. five words) can lead us to the same result for the

search process.

With respect to the data extraction process, if we

compare the topics generated from the 365 papers

(see Table 3) with queries 1 and 2, we identify three

matching words:

software

waterfall

, and

data

The topic model algorithm gives to each analyzed

paper a percentage of affinity with every topic. For

example, the paper S247 (Predicting data science

sociotechnical execution challenges by categorizing

data science projects) has 57% of affinity with Topic

1, 21% with Topic 2 and 22% with Topic 3.

The higher the percentage, the stronger the

relationship between the content of the article and the

topic is. As an example, Table 5 presents the

percentage of affinity for ten papers.

Table 5: Example of affinity between papers and topics.

Paper Topic 1 Topic 2 Topic 3

S241 21% 57% 22%

S242 19% 18% 63%

S243 20% 61% 19%

S244 19% 21% 60%

S245 57% 22% 21%

S246 57% 21% 22%

S247 57% 21% 22%

S248 18% 21% 61%

S249 18% 19% 63%

S250 20% 61% 19%

Next, we sort and give a ranking per topic for each

paper, where the higher percentage is ranking 1 and

the lower percentage is ranking 3, we group and count

how many papers belong to each group. For example,

the paper S245 corresponds to the group where topic

1 is ranked 1 (57%), topic 2 is 2 (22%), and topic 3 is

3 (21%). Others examples are the papers S244 and

S248 where the topic 1 is ranked 3, topic 2 is 2, and

topic 3 is 1.

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

282

Table 6 shows the summary of the percentage of

the 365 papers ranked per topic. The ranking position

1 represents which topic was assigned the highest

percentage. For example, the percentage of

documents whose ranking is 1 for topic 1 is 16%,

whereas the percentage of documents whose ranking

is 3 for topic 3 is 32%.

Table 6: Complete Topic Classification.

Ranking

1 2 3 Total

Topic

16% 44% 40% 100%

2 35% 37% 28% 100%

3 49% 20% 32% 100%

Total 100% 100% 100%

The topic classification for the 18 resulting papers

presented in Tables 1 and 2, is shown in Table 7. It

can be noticed that 44% of them belong to topic 1,

ranking 1.

Table 7: Topic classification of resulted papers.

Ranking

1 2 3 Total

Topic

44% 39% 17% 100%

2 28% 22% 50% 100%

3 28% 39% 33% 100%

Total 100% 100% 100%

From this sample, if we would like to use topic

modelling to reduce the number of articles to review,

we can say that reviewing only 16% of the total of

articles generated in the search, we could find 44% of

the results sought. In other words, we could get 8

(44% of 18) papers of our interest, avoiding reading

307 papers (i.e. only reading 58 out of 365 papers —

16% of the total number of articles).

3.5 Final Remarks

According to the sample presented in previous

section, we believe it is possible to use topic

modelling to reduce the number of articles to read,

filtering the papers more representatives to our

research. Although it is necessary to perform more

tests to improve the technique and increase the

percentage of success; the results presented here

demonstrate the benefits of using a topic mining

process.

Finally, with respect to the data analysis and

classification process, the manual SLR generates

information such as: methodologies used, Big Data

projects type, and the context or industry where they

were developed. However, this level of details was

not possible to obtain with a topic mining process. To

automate the classification process, we can try Fuzzy

techniques and supervised processes.

4 LIMITATIONS AND FUTURE

WORK

This article is limited to the search of Agile and Non-

agile Methodologies reported in case studies

associated with Big Data projects, excluding other

kinds of research such as formal experiments or

surveys. In addition, the searches were only executed

on four databases: Science Direct, Springer, Google

Scholar and Scopus. Also, the field of expertise of our

research team is mainly oriented to Software

Engineering.

As future work, we plan to replicate the whole

process with other kind of research studies to evaluate

how text mining contributes to the quality evaluation

of a SLR process and test Fuzzy techniques to

perform a supervised classification of the analysed

articles.

Additionally, we are interested in design a

framework for developing Big Data projects applying

agile principles.

5 CONCLUSIONS

The SLR carried out in this work demonstrates the use

of methodologies and process models since the

emergence of Big Data projects, increasing the use of

Agile Methodologies in this kind of projects from

2015 onwards. The methodologies most commonly

reported in publications related to Big Data projects

are: Scrum, XP, Kanban, and Crisp-DM.

According to the SLR, the applications of the Big

Data started in the scientific and academic fields

rather than the industrial and commercial sectors.

However, in the last two years, there has been an

increase in the number of Big Data projects in the

business field, especially in areas such as Marketing

and Innovation.

The integration of text mining as part of the

quality evaluation of the SLR process has allowed us

to test the ability of this technique to optimize this

kind of process.

REFERENCES

Al-Jaroodi, J., Hollein, B. and Mohamed, N. (2017).

Applying software engineering processes for big data

Application of Methodologies and Process Models in Big Data Projects

283

analytics applications development. In: 7th Annual

Computing and Communication Workshop and

Conference (CCWC). [online] Las Vegas: IEEE, pp. 1-

7. Available at: https://ieeexplore.ieee.org/document/

7868456 [Accessed 2 May 2018].

Batra, D. and Dahiya, R. (2016). Adapting Agile Practices

for Analytics Projects. In: 15th AIS SIGSAND

Symposium. [online] Lubbock: AIS, pp. 1-10. Available

at: https://pdfs.semanticscholar.org/96e5/

0da9309c6e2128d877e7f0842ac45d935410.pdf

[Accessed 7 Apr. 2017].

Campanelli, A. and Parreiras, F. (2015). Agile methods

tailoring – A systematic literature review. Journal of

Systems and Software, [online] 110, pp. 85-100.

Available at: https://www.sciencedirect.com/science/

article/pii/S0164121215001843 [Accessed 8 Nov.

2016].

Chen, H., Kazman, R., Garbajosa, J. and González, E.

(2016). Toward Big Data Value Engineering for

Innovation. In: 2016 IEEE/ACM 2nd International

Workshop on Big Data Software Engineering

(BIGDSE). [online] Austin: IEEE, pp. 44-50. Available

at: https://ieeexplore.ieee.org/document/7811386

[Accessed 5 Oct. 2017].

Chuang, J., Manning, C. D. and Heer, J. (2012, May).

Termite: Visualization techniques for assessing textual

topic models. In: Proceedings of the International

Working Conference on Advanced Visual Interfaces,

pp. 74-77. ACM.

Gandomi, A. and Haider, M. (2015). Beyond the hype: Big

data concepts, methods, and analytics. International

Journal of Information Management, [online] 35(2),

pp. 137-144. Available at: https://

www.sciencedirect.com/science/article/pii/S02684012

14001066 [Accessed 11 May 2016].

Kitchenham, B. and Charters, S. (2007), 'Guidelines for

performing Systematic Literature Reviews in Software

Engineering' (EBSE 2007-001), Technical report,

Keele University and Durham University Joint Report.

Kumar, V. (2017). Software Engineering for Big Data

Systems. Master Thesis. University of Waterloo.

Petersen, K., Vakkalanka, S. and Kuzniarz, L. (2015).

Guidelines for conducting systematic mapping studies

in software engineering: An update. Information and

Software Technology, [online] 64, pp. 1-18. Available

at: https://www.sciencedirect.com/science/article/pii/

S0950584915000646 [Accessed 5 Nov. 2016].

Wamba, S., Akter, S., Edwards, A., Chopin, G. and

Gnanzou, D. (2015). How ‘big data’ can make big

impact: Findings from a systematic review and a

longitudinal case study. International Journal of

Production Economics, [online] 165, pp. 234-246.

Available at: https://www.sciencedirect.com/science/

article/pii/S0925527314004253#! [Accessed 13 Oct.

2016].

APPENDICES

Appendices are available on: https://drive.

google.com/file/d/1ajQyGnUf0ONvPHjuosYcS6_tiu

Iu99MK/view?usp=sharing.

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

284