Content uploaded by Vusumuzi Maphosa
Author content
All content in this area was uploaded by Vusumuzi Maphosa on Oct 02, 2022
Content may be subject to copyright.
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5000
ARTIFICIAL INTELLIGENCE IN PROJECT MANAGEMENT
RESEARCH: A BIBLIOMETRIC ANALYSIS
1VUSUMUZI MAPHOSA, 2MFOWABO MAPHOSA
1Information Communication Technology and Services Department, National University of Science and
Technology, P.O AC 939, Ascot, Bulawayo, Zimbabwe
2School of Information Technology, Varsity College, Independent Institute of Education, Sandton, South
Africa
Email: 1vusumuzi.maphosa@nust.ac.zw, 2mmfowabo@varsitycollege.co.za
ABSTRACT
Projects are critical to organizations' success; hence improving project management (PM) is imperative.
Artificial intelligence (AI) has revolutionized many disciplines, including PM. Applying AI techniques in
PM can lead to more control for the project manager and better management of projects. One of the inherent
problems faced in PM is related to human error by the project managers leading to project failure. Project
managers use PM software and tools to improve their tasks. In this paper, we demonstrate the application of
AI in project management through a bibliometric analysis and keyword analysis to show the state of the art
of research on AI in PM in the past decade. We extracted 106 articles from the Web of Science database
published between 2012 and 2021 and analyzed the data. VOSviewer provided visual maps revealing
research hotspots in the field of AI in the HE knowledge base. Our analysis focuses on publication and
citation trends, the geographic distribution of articles, analysis of papers by source in which they were
published, h-index analysis, and keyword analysis. Results show that research in AI-based PM is widely
distributed geographically, by the publisher, and by discipline or field. Furthermore, research in the corpus
in the past decade has centered around four themes. The first theme relates to applying AI techniques to
improve accuracy, precision, and prediction in software projects, project management, and development
estimation. The second theme focuses on the application and development of AI for decision-making in PM.
The third theme highlights the benefits of applying AI in PM, such as dealing with uncertainty, improving
efficiency, scheduling, and stakeholder management. The last theme shows how AI manages risks and
improves cost management in engineering, procurement, and construction (EPC) projects. This research
makes valuable contributions to the corpus by highlighting opportunities, challenges, and future research
directions in AI in education. The study highlights its limitations and future research areas.
Keywords: Project Management, Artificial Intelligence, Bibliometric Analysis, Machine Learning
1. INTRODUCTION
Project management (PM) evolved from the
engineering and construction fields, and the past
decades have witnessed increased application
across various private and public practice fields.
A project is a temporary endeavor undertaken to
create a unique product, service, or result by
effectively organizing available resources to meet
the stated objectives1. Entities deliver complex
and advanced technological solutions, systems, or
services through projects. Through PM, entities
conceive ideas, develop them, and provide new
products to the customers. Projects play an
essential role in an organization and are vehicles
by which entities grow. Schoper et al.1 showed the
importance of projects in modern societies and
how the share of project-related work in advanced
economies is one-third and rising. Archibald and
Prado2 posit that projects generate a third of the
world's economy, and empirical evidence
supports the claim that most implemented projects
are unsuccessful.
The volatile nature of the 21st-century
economy and stiff competition force
organizations to complete projects within tighter
schedules and limited resources but still produce
top-quality outcomes3. Organizations must ensure
that PM aligns with the strategic goals and
efficient use of resources since projects play a
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5001
vital role in the business environment4. Projects
are critical to organizations' success; hence
improving PM is imperative4. Project
performance has been a subject of interest for a
long time5. Although projects have been
implemented for thousands of years, it was only
in the 20th century that PM became a recognized
discipline6.
Artificial intelligence (AI) refers to
developing "computing systems that can engage
in human-like processes such as learning,
adapting, synthesizing, self-correction and using
data for complex processing tasks." AI has
changed how jobs and tasks are performed,
bringing about automation and transformation. AI
techniques are being relentlessly applied in
numerous fields such as medicine, finance,
gaming, robotics, language understanding,
vehicle control, speech recognition, and planning
and scheduling. Shapiro and Eckroth7 state that
the first AI programs developed in the 1950s
could not solve complex problems, but they
enhanced the intelligent understanding of
problem-solving7. In the 1980s, the advancements
in AI systems made their use more cost-effective
for government and industry purposes. Experts
predict that by 2030, over 80% of repetitive and
mundane tasks representing the bulk of PM work
will be eliminated by collaboration between
humans and intelligent machines8. While AI is the
deployment of computing systems that can learn,
adapt, synthesize, self-correct and make
decisions. Applying AI in PM is not new. Based
on previous knowledge, Foster9 posited that AI
could be successfully applied in PM by analyzing
large datasets to find patterns, trends, and
problems that need attention.
Furthermore, AI could be used to monitor
how the project is going and make changes to
future activities if needed9. AI integration in PM
lies in the intersection of knowledge, learning,
adaptation, and decision making. PM and AI
integration will administer projects with a
minimum human intervention using smart
machines and large volumes of data to automate
decision-making, task management, and
prediction. With the help of AI, PM tasks can be
automated, and AI can direct a project and help
make related decisions.
Early scholars who studied AI application in
PM noted that AI could analyze large data sets to
discover trends and patterns and use that
information to support decision-making
processes10. More recently, Ong and Uddin11 note
a wide range of AI applications in projects that
reduce project risks, track project progress, and
identify irregularities and outliers in projects. AI
integration in PM can help estimate costs, handle
schedules, manage activities and follow-ups, and
plan resources12. With AI's help, PM tasks can be
done automatically, and AI can direct a project
and help make related decisions13. There is
evidence of research on using AI methods for
specific PM tasks. For example, AI can be used in
project forecasting, costing, and scheduling
success14. AI applications in PM can be
categorized into four types: autonomous
management of projects, Chabot assistance,
integration and automation, and machine-
learning-based15.
Bibliometric studies are different from
systematic literature reviews. Systematic
literature reviews examine the content of research
papers to identify the corpus's issues and patterns.
On the other hand, bibliometric studies involve
mapping out a corpus by exploring papers to
discover trends and uncover critical features,
patterns, and relationships of variables in a
particular field. De Bellis16 described
bibliometrics as quantitative methods that analyze
scientific literature using bibliometric data.
Descriptive data may be used in bibliometric
studies such as the author and the journal's
productivity index, collaborations, and citation
analysis.
Bibliometric analysis is a cross-disciplinary
science that uses statistical methods to analyze
knowledge and identify a particular domain's
development, growth, and future research
orientations17. Non-grey literature articles are
retrieved and analyzed through quantitative and
qualitative analysis to measure author distribution
patterns, keywords, citations, institutions, and
researchers' performance from a single database18.
However, analyzing bibliometric indicators and
literature mapping on multiple databases is often
challenging. Bibliometric analysis is a
methodological approach for collecting and
analyzing quantitative data concerning emerging
or consolidated research topics19. It is used in
different research fields, such as journal
performance, collaboration trends, and
uncovering emerging trends in articles and the
extant literature20.
The popularity of bibliometric analysis
studies can be attributed to the advancement,
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5002
availability, and accessibility of bibliometric
software such as CiteScore and VOSviewer,
scientific databases such as Scopus and Web of
Science, and the cross-disciplinary pollination of
the bibliometric methodology from information
science to business research20. By applying the
guidelines given by Donthu et al.20 on
bibliometric analysis, the research study was
developed to identify and analyze the following
performance analysis indicators :
The number of publications on AI in PM
published between 2012 and 2021;
The geographical publication trends in
the period under review;
The citations received by the selected
articles;
The distribution of articles by publisher
and discipline (field);
The research themes in the area.
Several studies on the application of AI in
different fields exist, for example – education
(Hinojo-Lucena et al.21), medicine (Guo et al.22,
Tran et al.23), operations (Dhamija and Bag24),
wastewater treatment (Zhao et al.25), higher
education (Maphosa and Maphosa26).
Furthermore, bibliometric studies have been done
in PM; for example, de Souza and Gomes26 and
Lechler and Yang27.
AI is likely to shape PM and play a significant
role in its development27. Nevertheless, there is no
bibliometric study related to AI application in PM
to our knowledge. The study aims to analyze
research on the application of AI in PM, realizing
the importance of artificial intelligence in
transforming most fields, such as PM. This study,
therefore, provides a baseline for researchers and
scientists to plan and research the integration of
these two fields. This study seeks to answer the
following questions:
What is the state of the art of research in
AI in PM?
Which countries have contributed the
most to the in the domain?
Which are the research hotspots and
emerging areas of research in the corpus?
The rest of this paper is structured as follows:
Section 2 outlines the methodology we followed,
how we collected data, and analyzed the data.
Section 3 presents the results using descriptive
methods and the outcomes from analysis done on
VOSViewer version 1.6.18. Section 4 discusses
the results' implications and reinforces our claims
about AI growth in PM. Finally, in Section 5, we
summarise the main findings of the bibliometric
analysis.
2. METHODOLOGY
We used bibliometric analysis, a scientific
computer-assisted review methodology, to
identify core domains or authors and map their
relationships by highlighting all the publications
related to a given topic or domain, thereby
enhancing the understanding of the overall
intellectual landscape16 36. We chose the Web of
Science database because it is the most extensive
citation and abstract database of peer-reviewed
publications. The Web of Science is a reputable
and widely used database that is easily accessible,
providing a basis for answering the research
questions posed in the previous sections. The Web
of Science has comprehensive search tools and
options that enable the researcher to apply several
Boolean operators, promoting an extensive search
query. Bibliometric data from the search query
can be analyzed and exported to other
bibliometric analysis software such as CiteSpace
and VOSViewer. For this study, we used
VOSViewer. In April 2022, we searched and
collected data with the initial search string
selecting articles focused on artificial intelligence
and project management. We searched articles
containing the terms "artificial intelligence" and
"project management" in the titles, abstracts, and
keywords. This search yielded 406 articles.
We then filtered the results to include articles
published between 2012 and 2021. One hundred
six articles were excluded leaving 300 articles.
We filtered by the article type, excluding
editorials and corrections, leaving 292 articles.
Finally, we filtered by language to include articles
published in English. Four articles were excluded
leaving 288 articles. The authors exported the
remaining 288 articles and the abstracts to M.S.
Excel for analysis. The analysis involved reading
the titles and abstracts for each of these articles.
Publications not related to artificial intelligence
and project management were excluded, leaving
106 articles that were analyzed. Figure 1 shows
the article selection process.
We downloaded data on 106 articles from the
Web of Science database in text format. We then
exported the abstract and keywords to
VOSviewer, a free bibliometric tool with
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5003
visualization and text-mining abilities. The Web
of Science database has embedded analytic
functions for citation, subject analysis, and h-
index analysis and supports the direct export of
data into VOSviewer and Excel. VOSviewer
provides network visualizations and density
maps28.
3. RESULTS
This study demonstrates the state of AI
application in PM, and the opportunities
contributed by integrating the two areas. This
study aims to comprehensively understand AI
applications in PM using bibliometrics to analyze
research conducted in the past decades. This study
explores the publication and citation trends, the
geographic distribution of publications, the
analysis of publications by top publishers, and the
leading research topics.
Table 1 shows the document types of the
articles retrieved. Journal articles (90.6%) are the
majority of the documents retrieved. Review
articles account for 9.4%. This section presents
the results of this study: the distribution of
publications by number published and citations
per year, the geographic distribution of articles,
distribution by publishers, analysis of
publications by field, and keyword analysis.
Table 1. Document types of the articles retrieved
Document Type
Count
Percentage
of 106
Articles
96
90.6
Review Articles
10
9.4
Early Access
4
3.8
Data Papers
1
0.9
Proceedings Papers
1
0.9
3.1 Distribution of Articles by Publications
and Citations
Figure 2 plots the publications and citation
trends. Between 2012 and 2019, less than ten
articles were published yearly, steadily increasing
from 2020 to 2021. In terms of citations, there
were no citations in 2012. 2013 and 2014 had four
and eight citations, respectively. 2015 to 2018 saw
a rise in citations, each year having more than 35
and less than 88. The last three years saw
astronomical growth in citations, with 2019
having 147, 2020 having 218, and 2021 having
420. Interest in AI research in PM has grown, with
2019, 2020, and 2021 has about 70% of all
publications. Similarly, citations have risen
astronomical, with the past three years accounting
for more than 77% of all citations in the decade.
The year 2021 accounted for nearly 42% of all
citations.
3.2 Geographic Distribution of Articles by
Country
It is essential to consider that for an item to be
accounted for in a country's production, at least
one of the authors must be associated with a
teaching or research institution in the related
country. Thus, the same scientific article may be
accounted for in more than one country29. The
articles' geographic distribution shows that 47
countries are represented. Table 2 gives the
geographic distribution of publications, and there
are currently three dominant countries: China
(18), the United States (17), and England (13).
Noteworthy is the absence of representation of
Africa among the top 15 most productive
countries. Africa has only three representatives
(Morocco, Egypt, and Tunisia).
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5004
Figure 1. Article Selection Process [30]
Figure 2. Trends Of Articles Published And Citations Per Year
534 4 6 6 4
7
18
49
048
38 36 51
87
147
218
420
0
10
20
30
40
50
60
0
50
100
150
200
250
300
350
400
450
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Publications
Citations
Publications Citations
Articles retrieved from initial search (n = 406)
Articles after refining search by years (n = 300)
Articles retrieved after refining by article type (n
= 292)
Articles retrieved after refining by language (n =
288)
Excluded editorial materials, corrections
and notes (n = 8)
Excluded articles not published
between 2012 -2021 (n = 106)
Excluded articles not published in
English (n = 4)
Excluded articles not covering both AI
and PM (n = 182)
Total articles included in the analysis (n = 106)
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5005
Researchers from Sub-Saharan Africa did not
contribute any articles written, which will
negatively impact the successful implementation
of projects requiring these skills.
Table 2. Articles Published Per Country
Rank
Country
Publications
1
China
18
2
USA
17
3
England
13
4
Canada
8
5
Spain
8
6
Taiwan
8
7
South Korea
7
8
India
6
9
Jordan
6
10
Australia
4
11
Vietnam
4
12
Belgium
3
13
France
3
14
Iran
3
15
Iraq
3
16
Malaysia
3
17
Morocco
3
18
Singapore
3
19
UAE
3
20
Cuba
2
21
Egypt
2
22
Germany
2
23
Greece
2
24
Pakistan
2
25
Poland
2
26
Portugal
2
27
Sweden
2
28
Austria
1
29
Brazil
1
30
Chile
1
31
Croatia
1
32
Cyprus
1
33
Finland
1
34
Hungary
1
35
Iceland
1
36
Italy
1
37
Japan
1
38
Lebanon
1
39
Mexico
1
40
New
Zealand
1
41
Norway
1
42
Romania
1
43
Saudi Arabia
1
44
Scotland
1
45
Serbia
1
46
Tunisia
1
47
Turkey
1
3.3 Distribution of Articles by Publisher
Thirty different publishers published 106
research articles. Table 3 shows the leading
publishers of research on AI and PM. Elsevier
dominated with 31 articles, followed by 'MDPI'
with 18 articles and then Springer Nature with 10.
IEEE has five publications, and the American
Society of Civil Engineers, Emerald Group
Publishing, Hindawi Publishing Group, and
Wiley have four. The distribution by publishers
shows that AI in PM is widespread across
different sectors.
Table 3. Leading publishers
Rank
Publisher
Publications
1
Elsevier
31
2
MDPI
18
3
Springer Nature
10
4
IEEE
5
5 American Society of
Civil Engineers
4
6 Emerald Group
Publishing
4
7 Hindawi Publishing
Group
4
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5006
8
Wiley
4
9
Atlantis Press
2
10
IOS Press
2
11
Taylor and
Francis
2
12 Vilnius Gediminas
Technical University
2
3.4 Analysis of Articles by Discipline
Table 4 shows the leading disciplines with
ten or more publications. Altogether, 37 fields or
disciplines are represented in the 106 articles
retrieved, indicating that PM is indeed a multi-
disciplinary field and AI's acceptance in the PM
field. Computer science - artificial intelligence
has the most significant publications, with 20
articles accounting for 18.8%, followed by civil
engineering, with 16 papers (15.1%), then
computer science information systems with 14
articles (13.1%). Computer science and
engineering are among the leading fields in
implementing AI in PM. This is expected, given
that AI is inherently a computer science subfield.
Table 4. Leading Disciplines
Rank
Field
Count
1 Computer Science Artificial
Intelligence
20
2
Engineering Civil
16
3 Computer Science Information
Systems
14
4 Construction Building
Technology
13
5 Computer Science Software
Engineering
12
6
Management
12
7 Engineering Electrical
Electronic
11
8
Engineering Multidisciplinary
10
9
Environmental Sciences
10
3.5 H-Index
The 106 articles were cited 1 042 times and
1 016 when excluding self-citations between 2012
and 2021. Each article has been cited with an
average of 11.08. The retrieved research papers
have an h-index of 21. The h-index of 21 means
that of the 106 research articles, 21 have received
at least 21 citations.
3.6 Keyword Analysis
This sub-section analyses the keywords used
in articles and abstracts using VOSviewer.
Keyword analysis has been a prominent research
theme in bibliometric studies and involves
counting the frequency of keywords to determine
hot spots within a particular research area28. To
highlight the topics AI addresses in PM, we
analyzed 3 310 terms in the titles and abstracts of
the 106 articles related to AI use in PM to
establish the co-occurrence network. Figure 3
shows the keyword co-occurrence network
visualization of high-frequency terms extracted
from titles and abstracts. Of the 3 310 terms
extracted, 124 had more than five occurrences,
and 74 met the co-occurrence threshold (top 60%
with the highest relevance). The link strength
between two items shows the co-occurrence
frequency – a quantitative depiction of the items'
relationship. The total link strength of an item is
the item's full link strength over all other items.
Table 5 shows the top 10 keywords in terms of
frequency. The table also shows the total link
strength of the keywords. The term 'accuracy' has
thicker lines with 'algorithm' (16), 'machine' (14),
'error' (11), and 'prediction' (9). In VOSviewer,
the stronger the link between two items, the
thicker the line used to display the link [28]. This
indicates a strong relationship between the terms'
problem' and 'algorithm,' perhaps pointing to the
fact that algorithms (a component of machine
learning and AI) are gaining prominence in
solving PM-related problems. The relationship
between 'algorithm' and 'machine,' 'development'
and 'industry' is quite strong.
Table 5: Top 10 Keywords And Their Total Link
Strength
Ran
k
Keyword
Occu
r
ren
ce
Link
s
Tot
al Link
Strength
1
Accuracy
32
71
265
2
Algorithm
31
69
230
3
Machine
29
64
220
4 Applicatio
n
21 61 160
5
Dataset
18
50
135
6 Developm
ent
17 54 140
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5007
7 Constructi
on
16 60 138
8
Prediction
16
51
107
9
Number
16
52
104
10
Type
15
53
108
Using VOSviewer, we generated four
clusters associated with the research topic from
the 74 terms. As shown in figure 3, the size of the
circle represents the frequency of the occurrence
of the term. The color of the circle represents the
cluster. The red and green clusters are the biggest;
each has 23 keywords, the blue cluster has 19, and
the yellow cluster is the smallest with 11. The red
cluster covers the research on AI-based PM for
improving accuracy, precision, and prediction in
software projects, project management, and
software development estimation. The cluster also
compares the performance of different algorithms
used in AI. The green cluster represents keywords
related to the application and development of AI
for decision-making in PM and practitioners and
researchers' role in adapting available technology
such as big data. The blue cluster highlights the
benefits of applying AI in PM, such as dealing
with uncertainty, improving efficiency,
scheduling, and stakeholder management. The
yellow cluster shows the use of AI in managing
risks and improving cost management in
engineering, procurement, and construction
(EPC) projects.
Figure 3. Keyword Co-Occurrence Network Visualization Of High-Frequency Terms
The keyword density visualization map is
shown in Figure 4. Density maps are valuable for
understanding the corpus organization and can
help draw attention to the map's most critical
areas32. In VOSviewer, colors indicate the density
of terms, ranging from green (lowest density) to
red (highest density)33. The red area keywords
appear more frequently, and the green color area
keywords appear less often. Two research
hotspots focused on accuracy, algorithm,
machine, prediction, neural networks, and support
vector machines. The other hotspot is centered
around application, development, and outcome.
Besides the hotspots, the map also shows that
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5008
research is growing to unexplored topics and areas
such as contract management, comparison of
different techniques, EPC, and big data.
4. DISCUSSION
This study aimed to conduct a bibliometric
analysis of the application of AI in PM. This study
contributes to the emerging literature on AI usage
in PM using bibliometric analysis. The
bibliometric analysis provides an insight into the
application of AI in PM. It shows how
innovatively AI is being applied in many non-
traditional applications to improve the success of
projects.
4.1 The state of the art of research in the field
of AI in PM
Project success is essential in the world
economy; it is crucial to understand research
applying AI in PM. There is a noticeable increase
in the number of publications in 2020 and 2021
and citations between 2019 and 2021. This
increase indicates the general acceptance of AI
tools and PM techniques, especially in the past
decade. Given the increasing number of citations
for articles on AI application in PM, we expect
research in this area to increase.
Figure 4. Keyword Density Visualization Map
With 106 articles in AI-based PM
published in the past decade is an indication that this
research domain is yet to mature. The spike in
publications between 2019 and 2021 shows that the
field is growing. AI is poised to transform the field
of project management significantly. The wide range
of publishers publishing articles on AI in PM shows
the interest that the field of AI in PM is receiving,
indicating its widespread acceptance across the
different fields.
Analysis of the publishers reveals that
leading publishers are publishing research. Over
90% of the articles on AI in PM are journal type and
were published in the past decade. This demonstrates
the quality of the research being done as journals
have rigorous and stringent requirements for
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5009
publishing. Furthermore, the quality of these articles
is evidenced by the h-index of 21 and the citations
the articles have received.
4.2 Research by countries
Advanced computer applications have
allowed applications such as AI to permeate the
industrial, services, and manufacturing sectors'
ecosystems through innovative solutions. China and
the USA lead in AI in PM research, the two most
industrialized and the world's largest economies.
There is very little research on Africa and the Middle
East. Computer Science and Engineering disciplines
dominate studies on the application of AI in PM.
Overall, this study's findings indicate that AI
techniques in PM are still in infancy but growing.
This growth is shown by the increasing number of
publications and citations in other research articles.
4.3 Research hotspots and emerging areas
Our results show that neural networks,
algorithms, big data, and machine learning are
popular AI techniques used in project management.
Additionally, the resources used to proffer
technological solutions. PM is an applied field, and
this is shown by the wide variety of fields that
published articles about the application of AI in PM.
The two dominant fields are computer science and
engineering. This is expected as AI is a subfield of
computer science.
One would expect distinct subfields to
emerge as the application of AI grows. Ong and
Uddin (2020) noted that the current AI and data
science applications remain preliminary with room
for improvement. In PM, there is a need for software
and tools that are proactive and not reactive. These
need to predict future scenarios using 'what if'
analysis and alert the project manager before issues
arise33. This will undoubtedly change the role of a
project manager as we know it today. Currently, the
usage of AI tools in PM by project managers is low,
according to a study by the International Project
Management Association (IPMA) and
PricewaterhouseCoopers (PwC)34.
There is an increase in incorporating
chatbots and machine learning techniques in project
management. Adopting AI in project management
assists project managers in effectively and
efficiently managing their daily tasks. AI-based
project management also improves accuracy,
insight, decision-making, and strategy, thus
improving productivity. Machine learning
techniques use predictive and corrective analysis
techniques in project planning and management35.
Expert systems can assist project managers in
grasping experiences of managing previous projects
to share experiences easily. Project managers face
challenges in predicting future scenarios and
proactively taking corrective measures, and
predictive analytics improves predicting 'what if'
developments and uncovering real-time insights.
Some of the failures in PM are a direct result of
human limitations. The use of AI can reduce the
mistakes project managers make. According to Price
Water House Coopers [34], AI is one of the
breakthroughs that will disrupt PM by creating AI-
enabled PM tools that enhance decision making. We
expect an increase in project success rates as PM and
AI become increasingly integrated. Critical to this
development is the role of research.
5. LIMITATIONS
Although great effort went into conducting this
study, there are inherent shortcomings. One
shortcoming emanates from using one database
(Web of Science), meaning any limitations in the
database may apply to the study in turn. Thus,
valuable grey literature and publications not indexed
to the Web of Science have been excluded. This
results in the underestimating research trends in most
developing countries whose authors cannot publish
in top journals indexed with the database due to their
rigor and high standards. Each document is counted
once per author in the Web of Science database.
There is an overlap when counting authors with
different affiliations, possibly over-estimated active
authors and countries. Thus countries with authors
involved in collaboration research are counted even
if the author is a co-author and not the principal
author. This may misrepresent patterns, trends, and
patterns in the field. Another limitation stems from
filtering grey literature, such as articles relating to
reviews, letters, and patents in the study.
6. CONCLUSION
To the authors' knowledge, this is the first
bibliometric analysis of AI in PM research by
scholars from sub-Saharan Africa. This study's
findings provide insight into AI in PM and show a
sharp rise in research articles published in the last
three years. The data analysis from the Web of
Science database provided insights into the
distribution of the retrieved articles by year,
geographic region, publisher, and field. The results
show a widespread uptake of AI tools in PM across
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5010
different countries and diverse fields. The keyword
analysis determined the research topics and trends in
AI applications in PM.
Our research contributes to the application of AI
in PM and encourages further research on this
important topic for the future. There is a need to
evaluate how advances in quantum computing and
cognitive technologies can be employed in project
management. The study findings suggest that there
are still research gaps in the field. One dominant
research subfield relating to applying AI techniques
to solve PM's inherent problems can be seen clearly
in the density visualization map (figure 4) –
algorithms to improve accuracy, prediction, and
machine learning techniques. One distinct subfield's
presence is a perhaps justifiable observation given
the relative newness of AI applications in PM. This
finding elucidates a need for further research in the
less distinct subfields, including but not limited to
AI's impact on PM in these fields. Future research
should evaluate the project managers' knowledge of
AI and how they are prepared to work with AI.
The world is experiencing a sharp increase in
applying new technologies in the industrial and
services sectors. While this study has provided an
overview of AI applications in PM, there is a need to
quantitatively measure the impact of AI application
usage on general project success. This study
provides literature on the application of AI in project
management and an understanding of the critical
relationships between the two fields. Our study
aimed to stimulate and attract more research interest
in AI application in PM. This study provides
valuable insight for researchers and policymakers in
PM and AI
REFERENCES:
[1]
Y.
Schoper, A. Wald, H. Ingason and T.
Fridgeirsson, "Projectification in Western
economies: A comparative study of Germany,
Norway,
and Iceland," International Journal of
Project Management, vol. 36, no. 1, pp. 71
-
82, 2018
.
[2]
R. D. Archibald and D. Prado, "
The importance of
knowing your project, programme and portfolio
management maturity,"
Project Management World
Journal,
vol. 3, no. 11, 2014
.
[3]
R. Wagner, "
Organisational competence in project
management —
new perspectives on assessing and
developing organisations,
"
Journal of Project,
Programme and Portfolio Management,
vol. 3, no.
1, 2012
, pp. 45
-
57
.
[4]
M. Mullaly, "If maturity is the answer, then exactly
what was the question?
" International Journal of
Managing Projects in
Business, vol. 7, no. 2, pp.
169
-
185, 2014
.
[5]
G. Garel, "A history of project management models:
From pre-
models to the standard models,"
International Journal of Project Management, vol.
31, pp. 663
-
669, 2013.
[6]
M. Voss, "Impact of customer inte
gration on project
portfolio management and its success —
Developing a conceptual framework," International
Journal of Project Management, vol. 30, pp. 567-
581, 2012
.
[7]
S. C. Shapiro and D. Eckroth, Encyclopedia of
Artificial Intelligence, New York: W
iley, 1987
.
[8]
R. van der Meulen, "Smart Machines To Take Over
Most Routine PPM Work," 2017. [Online].
Available:
https://www.gartner.com/smarterwithgartner/smart-
machines-to-take-over-most-routine-ppm-work/.
[Accessed 01 February 2022].
[9]
A.
Foster, "Artificial intelligence in project
management," Cost Engineering, 30(6), 21., vol. 30,
no. 6, pp. 21
-
30, 1988.
[10]
R. Levitt and J. Kunz, "Using artificial intelligence
techniques to support project management," AI
EDAM, vol. 1, no. 1, pp. 3
-
24, 1987.
[11]
S. Ong and S. Uddin, "
Data science and artificial
intelligence in project management: The past,
present, and future,"
The Journal of Modern Project
Management,
vol. 7, no. 4, 2020
, pp. 1
-
9
.
[12]
M. Somasundaram, K. Junaid and S. Mangadu
,
"
Artificial Intelligence (AI) Enabled Intelligent
Quality Management System (IQMS) For
Personalized Learning Path,"
Procedia Comput.
Sci.,
vol. 172, 2020
, pp. 438
–
442
.
[13]
M. Munir, "
How artificial intelligence can help
project managers,"
Global Journal of Management
and Business Research, vol. 19, no. 4, 2019, pp. 29-
35.
[14]
M. Wauters and M. Vanhoucke, "A comparative
study of Artificial Intelligence methods for project
duration forecasting," Expert systems with
applications, vol. 46, p
p. 249
-
261, 2016.
[15]
Q. Wang, "
How to apply AI technology in project
management,"
PM World Journal, vol. 8, no. 3,
2019, pp. 1
-
12.
[16]
N. De Bellis, Bibliometrics and Citation Analysis:
From the Science Citation Index to Cybermetrics,
Plymouth: The Scarecrow Press Inc, 2009.
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5011
[17]
J. Merigó, "Academic research in innovation: A
country analysis. Scientometrics," vol. 108, p. 559–
593, 2016.
[18]
A. Van Raan, "Advances in bibliometric analysis:
research performance assessment and scienc
e
mapping," Bibliometrics Use and Abuse in the
Review of Research Performance, p. 17
–
28, 2014.
[19]
G. Fusco, "Twenty years of common agricultural
policy in Europe: A bibliometric analysis,"
Sustainability, vol. 13, no. 19, pp. 1
-
15, 2021.
[20]
S. Don
thu, "How to conduct a bibliometric analysis:
An overview and guidelines," Journal of Business
Research, vol. 133, pp. 285
-
296, 2021.
[21]
F. J. Hinojo-Lucena, I. Aznar-Díaz, M. P. Cáceres-
Reche and J. M. Romero-
Rodríguez, "Artificial
intelligence in higher education: A bibliometric
study on its impact in the scientific literature,"
Education Sciences, vol. 9, no. 1, p. 51, 2019
.
[22]
Y. Guo, Z. Hao, S. Zhao, J. Gong,
and F. Yang,
"Artificial intelligence in health care: Bibliometric
analysis," Journal of Medical Internet Research, vol.
22, no. 7, p. e18228, 2020
[23]
B. X. Tran, R. S. McIntyre, C. A. Latkin, H. T. Phan,
G. T. Vu, H. L. Nguyen, K. K. Gwee, C. S. Ho
and
R. Ho, "
The current research landscape on the
artificial intelligence application in the management
of depressive disorders: A bibliometric analysis,"
International Journal of Environmental Research
and Public Health, vol. 16, no. 12, 2019, pp. 2150-
21
60.
[24]
P. Dhamija and S. Bag, "Role of artificial
intelligence in operations environment: A review
and bibliometric analysis," The TQM Journal, 2020
.
[25]
L. Zhao, T. Dai, Z. Qiao, P. Sun , J. Hao and Y.
Yang, "Application of artificial intelligence
to
wastewater treatment: A bibliometric analysis and
systematic review of technology, economy,
management, and wastewater reuse," Process Safety
and Environmental Protection, vol. 133, pp. 169-
182, 2020.
[26]
V. Maphosa, M. Maphosa, "
The Trajectory of
A
rtificial Intelligence Research in Higher
Education: A Bibliometric Analysis and
Visualisation,"
in International Conference on
Artificial Intelligence, Big Data, Computing and
Data Communication Systems (icABCD), Durban,
South Africa, 2021, pp. 1
-
7.
[27]
T. Fridgeirsson, H. Ingason, H. Jonasson and H.
Jonsdottir, "An authoritative study on the near future
effect of artificial intelligence on project
management knowledge areas," Sustainability, vol.
13, no. 4, pp. 2345
-
2354, 2021.
[28]
N. J. van
Eck and L. Waltman, "VOSviewer manual:
Manual for VOSviewer version 1 .6.11," Leiden
University, Leiden, 2017
[29]
R. Côrtes, "A bibliometric study on "
education for
sustainability,"
Brazilian Journal of Science and
Technology, vol. 3, no. 1, pp. 1
-
17, 2
016.
[30]
A.A Selçuk, "
A Guide for Systematic Reviews:
PRISMA,"
Turk Arch Otorhinolaryngol, vol. 57, no.
1, 2019, pp. 57
-
8.
[31]
C. Huai and L. Chai, "A bibliometric analysis on the
performance and underlying dynamic patterns of
water security research,
" Scientometrics, vol. 108,
no. 3, pp. 1531
-
1551, 2016.
[32]
N. V. Chawla and D. A. Davis, "
Bringing big data
to personalised healthcare: A patient-
centered
framework,"
Journal of General Internal Medicine,
vol. 28, no. 3, 2013
, pp. 660
-
665
.
[33]
T.
G. Lechler and S. Yang, "Exploring the role of
project management in the development of the
academic agile software discourse: A bibliometric
analysis," Project Management Journal, vol. 48, no.
1, pp. 3
-
18, 2017.
[34]
IPMA and PwC, "Artificial Intelligenc
e impact in
Project Management,
"
IPMA and PwC, 2020.
[35
]
R. Prieto, "Impacts of Artificial Intelligence on
Management of Large Complex Projects," PM
World Journal, vol. 8, no. 5, pp. 1
-
20, 2019.
[36]
V. Maphosa, "
Rethinking Sustainability: A
Bibliometric and Visualisation of e-
waste
Management in Africa, "Journal of Highe
r
Education Theory and Practice, vol 22, no. 1
, pp
123
-
135, 2022
.
Journal of Theoretical and Applied Information Technology
31st August 2022. Vol.100. No 16
© 2022 Little Lion Scientific
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
5012
Appendix A. Steps for Analysing using
VOSviewer
1. Download VOSviewer
(https://www.vosviewer.com/download)
2. Install VOSviewer
3. Download the bibliometric data in plain
text from the Web of Science database or
another source and save it in a single folder.
4. Start VOSviewer, click the "Create" button,
and select the second option, "Create a map
based on bibliographic data," on the chosen
data type.
5. Click the "Next" button and select the first
option, "Read data from bibliographic
database files," on the chosen data source.
6. Click the 'Next" button, select the file step,
and select the file downloaded from the
database.
7. Click the "Next" button, and on the next
step, "Choose the type of analysis and
counting method step," select type of
analysis – Co-occurrence, Unit of Analysis
- All keywords and Counting Method: - full
counting.
8. Click finish
