Bibliometrische visualisaties voor het bijhouden van
wetenschappelijke literatuur
Nees Jan van Eck
Centre for Science and Technology Studies (CWTS), Leiden University
VOGIN-IP-lezing
21 oktober 2021

About myself
• Master in computer science
• PhD thesis on bibliometric mapping of science
• Senior researcher at CWTS
– Bibliometric network analysis and visualization
– Bibliometric data sources
– Bibliometric indicators
• Head of ICT
1

Centre for Science and Technology Studies (CWTS)
• Research center at Leiden University
• Science and technology studies, with a considerable
emphasis on scientometrics
• About 50 staff members
• Our mission: Making the science system better!
• Full access to large bibliographic database (Web of
Science, Scopus, Dimensions, PubMed, PATSTAT,
etc.)
2

Centre for Science and Technology Studies (CWTS)
• Basic research:
– Quantitative science studies
– Science and evaluation studies
– Science, technology and innovation studies
• Commissioned research:
– Bibliometric studies for universities, funding organizations, governments, scientific publishers, etc.
– Mostly done using the in-house databases of CWTS
3

Program
• Introduction
• Hands-on assignments
– Data collection
– VOSviewer:
• Co-authorship maps
• Citation-based maps
• Term maps
4

Introduction
5

Explosive growth of scientific literature
6
“A new scientific paper is published
every 30 seconds and there are
10,000 updates to PubMed every day”
“It is impossible for humans alone to
process all of the complex information
potentially available to them for the
advancement of scientific research”
James Chandler, November 2016, BenevolentAI

Bibliographic data sources
7

Bibliographic data sources
8
Web of Science Scopus Dimensions
Journals 20,000 25,000 50,000
Publications 60 million 50 million 110 million
Citations 1.4 billion 1.5 billion 1.2 billion

Bibliometric networks
9
WoS
Scopus
Dimensions
Lens
MAG
Semantic Scholar
PubMed
Europe PMC
Crossref
Citation network
of pubs / journals / authors / orgs / countries
Co-authorship network
of authors / orgs / countries
Co-citation network
of pubs / journals / authors / orgs / countries
Co-occurrence network
of keywords / terms
Bibliographic coupling network
of pubs / journals / authors
Bibliographic
data source

VOSviewer
10

VOSviewer
• Focus on visualization of scientometric networks
• Support for large number of data sources
• Text mining functionality
• Advanced visualization features
• Relatively easy to use
• Limited analysis options
• Developed at CWTS
11

Users of VOSviewer
• Researchers
• Research institutions
• Research funders
• Scientific publishers
• Industry
12

13

14

Text mining
15
JOI
Scientometrics
JASIST

Development of VOSviewer
16

Types of networks supported by each data source
17
Co-authorship Co-occurrence Citation Bibliographic
coupling
Co-citation
Web of Science     
Scopus     
Dimensions     
PubMed     
Crossref     
Europe PMC     
Lens     
Microsoft Academic     
Semantic Scholar     
OCC / COCI     
Wikidata     

Data access
18
Download

Interactive querying
19

20

21

22

Data collection
23

Data access
24
Download

Downloading Web of Science data
Pay attention to the following:
• Use the Web of Science Core Collection database
• Use the Full Record and Cited References option
• Use a suitable file format:
– VOSviewer supports both the plain text and the tab-delimited format
– Tab-delimited files can be easily processed using spreadsheet software
• We use the tab-delimited format
25

26
Downloading Web of Science data

Downloading Web of Science data
27

Downloading Web of Science data
28

Downloading Scopus data
• Use the CSV export option, not the Download option!
29

Downloading Scopus data
30

Downloading Scopus data
31

Downloading Scopus data
32

Downloading Dimensions data
• Make sure to create a user account!
• Extract the CSV file from the downloaded ZIP file
33

Downloading Dimensions data
34

Downloading Dimensions data
35

Downloading Dimensions data
36

Downloading Lens data
• Create a user account if you want to export more than 1,000 search results
37

Downloading Lens data
38

Downloading Lens data
39

Demonstration + Data collection assignment
40

Co-authorship maps
41

Co-authorship map of researchers
42

Co-authorship map of organizations
43

Co-authorship map of ASMS members
44

Co-authorship map of research organizations
45

Demonstration + Co-authorship map assignments
46

Citation-based maps
47

Determining relatedness based on citation data
• Determining relatedness of publications, journals, authors, etc. based on
citation data:
– Direct citation relations
– Bibliographic coupling relations
– Co-citation relations
• VOSviewer supports all three types of citation-based relations
• Another tool, CitNetExplorer, is focused entirely on direct citation relations
48

Direct citation relations
49
P1 P2
P1 and P2 are related because P1 has cited P2
(i.e., P2 appears in the reference list of P1)

Bibliographic coupling relations
50
P2
P3
P4
P1
Citing papers
Cited papers
P1 and P2 are related
because they have cited
P3, P4, and P5 in their
reference list
P5

Co-citation relations
51
P2
P3
P1
Citing papers
Cited papers
P1 and P2 are related
because they are co-cited
in the reference lists of P3,
P4, and P5
P4
P5

Direct citation network of publications
52
P1
P2
P3
P4
P5
P6
Citation network
(directed)
Citation network
(undirected)
P1
P2
P3
P4
P5
P6

Bibliographic coupling network of publications
53
P1
P2
P3
P4
P5
P6
P1
P2
P3
P4
P5
P6
2
1
1
1
Citation network
Bibliographic
coupling network

Co-citation network of publications
54
P1
P2
P3
P4
P5
P6
P1
P2
P3
P4
P5
P6
2
1
1
1
Citation network
Co-citation network

Comparison of different types of citation-based
relations
• Direct citation relations:
– Focus on recent publications
– Stable (direct citation relations won’t change in the future)
– Some publications may have no direct citation relations
• Bibliographic coupling relations:
– Focus on recent publications
– Stable (bib. coupling relations won’t change in the future)
– Publications almost always have bib. coupling relations
• Co-citation relations:
– Focus on older publications
– Unstable (co-citation relations may change in the future)
– Some publications may have no co-citation relations
55

Journal citation map
56

Journal co-citation map

Interpretation of a citation-based map
• Size:
– The larger an item (e.g., author or journal), the larger the number of publications of the item
• Distance:
– In general, the smaller the distance between two items, the higher the relatedness of the items, as
measured by citations, bibliographic coupling or co-citations
– The horizontal and vertical axes have no special meaning; maps can be freely rotated and flipped
• Colors:
– Colors indicate clusters of strongly related items
58

Citation-based maps supported by VOSviewer
• Citation and bibliographic coupling maps:
– Publications
– Journals
– Authors
– Organizations
– Countries
• Co-citation maps:
– Cited references
– Cited journals
– Cited authors (first author only in the case of Web of Science)
59

Demonstration + Citation-based maps assignments
60

Term maps
61

Term co-occurrence network

Interpretation of a term map
• Size:
– The larger a term, the higher the frequency of occurrence of the term
• Distance:
– In general, the smaller the distance between two terms, the higher the relatedness of the terms, as
measured by co-occurrences
– The horizontal and vertical axes have no special meaning; maps can be freely rotated and flipped
• Colors:
– Colors indicate clusters of closely related terms
63

Term co-occurrence network: Journal overlays

Term co-occurrence network: Time trend overlay

Term co-occurrence network: Citation impact overlay

Term map of Dutch biomedical research
67

Term map of Dutch biomedical research: Activity
overlay LUMC
68

Term map of Dutch biomedical research: Activity
overlay VUmc
69

Term map of Alzheimer research: Overlay based on
publication year
70

Term map of clinical neurology research: Citation
overlay
71

Demonstration + term map assignments
72

Wrap-up
73

Aim of bibliometric maps
• To provide an overview of the structure of the scientific literature in a certain
domain or on a certain topic
• Applications:
– To identify the main research areas within a scientific field
– To get insight into the size of the different areas
– To see how the areas relate to each other
• Bibliometric maps are especially useful when one has to deal with a relatively
large body of literature
74

Bibliometric mapping process
75
research
question
bibliographicd
ata
bibliometric
map
results
data collection data processing
analysis
interpretation
revision

Use of scientometric visualizations
• Visualizations can be used to make complex information more accessible
• Visualizations need to be used in combination with expert knowledge:
– Confirm expert knowledge
– Challenge expert knowledge
– Draw expert’s attention to interesting information
• Be aware of information loss and methodological sensitivity
• Visualizations can also be used pragmatically to attract attention and to ‘sell
a message’
76

Do it yourself!
77
www.vosviewer.com www.citnetexplorer.nl

Course: Visualizing Science Using VOSviewer
• January 25-28, 2022
• 4 half-day online sessions
• Participants are introduced into the main techniques for visualizing science
• Special attention is paid to applications in a research evaluation and
research management context
• The course has an interactive format and the small group size allows
participants to practice extensively with the VOSviewer software
• www.cwts.nl
78

Thank you for your attention!
79

Further reading
80

Further reading
81