VOSviewer Manual

Nees Jan van Eck and Ludo Waltman

28 September 2016

Manual for VOSviewer version 1.6.5

Table of contents
1 Introduction ........................................................................................... 2

2 User interface ........................................................................................ 3

2.1 Main panel ...................................................................................... 4

2.1.1 Network visualization ................................................................. 4

2.1.2 Overlay visualization.................................................................. 5

2.1.3 Density visualization .................................................................. 6

2.1.4 Zooming and scrolling ................................................................ 8

2.2 Options panel .................................................................................. 9

2.3 Information panel .......................................................................... 12

2.4 Overview panel .............................................................................. 12

2.5 Action panel .................................................................................. 12

2.5.1 File tab .................................................................................. 13

2.5.2 Items tab ............................................................................... 15

2.5.3 Analysis tab ............................................................................ 16

3 File types............................................................................................. 20

3.1 Map file ........................................................................................ 20

3.2 Network file .................................................................................. 22

3.3 Thesaurus file ................................................................................ 22

3.4 Cluster colors file, overlay colors file, and density colors file ................ 23

4 Advanced topics ................................................................................... 25

4.1 Using command line parameters ...................................................... 25

4.2 Making a map available online ......................................................... 29

4.3 Increasing the availability of memory ............................................... 30

References ................................................................................................. 31

1
1 Introduction
VOSviewer is a computer program for creating maps based on network data and
for visualizing and exploring these maps. The main features of VOSviewer can be
summarized as follows:

Creating maps based on network data. Maps can be created based directly on a
network, but it is also possible to create maps of scientific publications,
scientific journals, researchers, research organizations, countries, or keywords
based on co-authorship, co-occurrence, citation, bibliographic coupling, or co-
citation networks extracted from Web of Science, Scopus, PubMed, or RIS files.
Term maps can be created directly based on a text corpus. Maps are created
using the VOS layout technique and the VOS clustering technique.

Visualizing and exploring maps. Three visualizations are provided, the network
visualization, the overlay visualization, and the density visualization. Zooming
and scrolling functionality allows maps to be explored in full detail, which is
essential when working with large maps containing hundreds or even thousands
of items.

Although VOSviewer is intended primarily for analyzing bibliometric networks, the

program can in fact be used to create, visualize, and explore maps based on any
type of network data.

VOSviewer is written in the Java programming language, which means that it runs
on most hardware and operating system platforms. VOSviewer can be obtained
from www.vosviewer.com. The program can be used freely for any purpose.

This manual is concerned with version 1.6.5 of VOSviewer. The manual is

organized as follows. We first introduce the user interface of VOSviewer in Chapter
2, we then explain the file types used by VOSviewer in Chapter 3, and finally we
discuss some advanced topics in Chapter 4. For additional information about
VOSviewer, we refer to a paper that we have written (Van Eck & Waltman, 2010).
In this paper, a general introduction to VOSviewer is provided. Also, the technical
implementation of specific parts of the program is discussed in considerable detail.
Similar information, including a step-by-step tutorial, can also be found in a more
recent book chapter (Van Eck & Waltman, 2014).

2
2 User interface
The main window of VOSviewer is shown in Figure 1. As can be seen in the figure,
the main window consists of the following five panels:

Main panel. In this panel, a visualization of the currently active map is

presented. Zoom and scroll functionality can be used to determine the area in
the currently active map that is shown in the main panel.

Options panel. This panel can be used to change the way in which the currently
active map is visualized in the main panel.

Information panel. In this panel, the description of an item in the currently

active map can be shown.

Overview panel. In this panel, an overview of the currently active map is

presented. A rectangular frame is displayed in the overview panel to indicate
the area in the currently active map that is shown in the main panel.

Action panel. This panel can be used to undertake all kinds of actions, such as
creating a new map, opening or saving an existing map, making a screenshot,
and updating the layout or the clustering of a map.

VOSviewer provides three visualizations, which are referred to as the network

visualization, the overlay visualization, and the density visualization. As shown in
Figure 1, the Network Visualization, Overlay Visualization, and Density
Visualization tabs can be used to switch between the three visualizations.

In the next sections, we discuss the five panels of VOSviewer in more detail.

To get some hands-on experience with VOSviewer, we encourage the reader to use
the map file journal_map.txt and the network file journal_network_sparse.txt,
which are both distributed together with VOSviewer. The files contain a map of a
co-citation network of 232 journals in the fields of economics, management, and
operations research (for more details, see Van Eck & Waltman, 2010). To open the
map, click the Open button on the File tab in the action panel, select the map file
journal_map.txt and the network file journal_network_sparse.txt, and click
the OK button.

3
 4
4

1
5 2
4

Figure 1. The main window of VOSviewer. The numbers indicate

(1) the main panel, (2) the options panel, (3) the information panel,
(4) the overview panel, and (5) the action panel.

2.1 Main panel

As can be seen in Figure 1, the main panel is used to present a visualization of the
currently active map. Zoom and scroll functionality can be used to determine the
area in the currently active map that is shown in the main panel. The way in which
the currently active map is visualized depends on whether the network
visualization, the overlay visualization, or the density visualization is selected.

2.1.1 Network visualization

When the network visualization is selected, items are indicated by their label and,
by default, also by a circle. For each item, the size of the items label and the size
of the items circle depend on the weight of the item. 1 The color of an item is
determined by the cluster to which the item belongs. We note that for some items

1
The weight of an item can be determined by the weight (or normalized weight) column in a map file
(see Section 3.1). When a network is available, two weights are provided automatically. One weight is
the number of links of an item. The other weight is the total strength of the links of an item.

4
the label may not be visible. This is done in order to avoid overlapping labels. Also,
by default, no lines between items are displayed. However, this can be changed by
increasing the number of lines in the Lines text box in the options panel.

An example of the network visualization is shown in Figure 2. The distance

between two journals in the visualization approximately indicates the relatedness
of the journals in the co-citation network. In general, the closer two journals are
located to each other, the stronger their relatedness in terms of co-citation links.
Lines indicate the strongest co-citation links between journals.

Figure 2. The network visualization.

2.1.2 Overlay visualization

The overlay visualization is identical to the network visualization except that items
are colored differently. There are two ways in which items can be colored in the
overlay visualization. If scores have been given to items (using the score column in
a map file; see Section 3.1), the color of an item is determined by the items score,
where by default colors range from blue (low score) to green (average score) to
red (high score). On the other hand, if user-defined colors have been specified for
items (using the red, green, and blue columns in a map file; see Section 3.1), the

5
color of an item is determined by the items user-defined color. If neither scores
nor user-defined colors are available, the overlay visualization is not available.

An example of the overlay visualization is shown in Figure 3. A color bar is shown

in the lower right corner of the visualization. This color bar is shown only if colors
are determined by the scores of items. The color bar indicates the correspondence
between colors and scores. In the overlay visualization shown in Figure 3, colors
indicate the impact factors of journals.

Figure 3. The overlay visualization.

2.1.3 Density visualization

There are two variants of the density visualization. We first discuss the item
density visualization, followed by the cluster density visualization. The Item
density and Cluster density radio buttons in the options panel can be used to
switch between the two variants of the density visualization. We refer to Van Eck
and Waltman (2010) for a detailed discussion of the technical implementation of
the density visualization.

In the item density visualization, items are indicated by their label in a similar way
as in the network visualization and the overlay visualization. Each point in a map

6
has a color that depends on the density of items at that point. By default, this color
is somewhere in between red and blue. The larger the number of items in the
neighborhood of a point and the higher the weights of the neighboring items, the
closer the color of the point is to red. Conversely, the smaller the number of items
in the neighborhood of a point and the lower the weights of the neighboring items,
the closer the color of the point is to blue. An example of the item density
visualization is shown in Figure 4.

Figure 4. The item density visualization.

The cluster density visualization is available only if items have been assigned to
clusters. The cluster density visualization is similar to the item density visualization
except that the density of items is displayed separately for each cluster of items. In
the cluster density visualization, the color of a point in a map is close to the color
of a certain cluster if there are a large number of items belonging to that cluster in
the neighborhood of the point. Like in the item density visualization, items with
high weights count more heavily than items with low weights. An example of the
cluster density visualization is shown in Figure 5.

7
 Figure 5. The cluster density visualization.

2.1.4 Zooming and scrolling

To facilitate the detailed examination of a map, VOSviewer offers zoom and scroll
functionality. In the main panel, zooming and scrolling can be done in the following
three ways:

Using the mouse. To zoom in, move the mouse upwards while keeping the right
mouse button pressed. Conversely, to zoom out, move the mouse downwards
while keeping the right mouse button pressed. As an alternative, the mouse
wheel can be used to zoom in and out. To scroll through a map, move the
mouse while keeping the left mouse button pressed.

Using the navigation buttons in the upper right corner of the main panel (see
Figure 1). Use the plus and minus buttons to zoom in and out. Use the arrow
buttons to scroll through a map.

Using the keyboard. Use the plus and minus keys to zoom in and out. Use the
arrow keys to scroll through a map.

8
2.2 Options panel
The options panel can be used to change the way in which the currently active map
is visualized in the main panel. The network visualization, the overlay visualization,
and the density visualization each provide somewhat different options. Some of the
options are not always available. The options panel shows only the options that are
relevant given the currently available map data.

When the network visualization or the overlay visualization is selected, the

following options may be available:

Visualization.

o Scale. This slider determines the size of the labels of items in the main
panel. The Scale slider also determines the thickness of the lines used to
display links between items.

o Weights. When items have multiple weights, the Weights drop down list
determines the weights that control the size of the labels of items in the
main panel.

o Scores. This drop down list is available only if the overlay visualization is
selected. When items have multiple scores, the Scores drop down list
determines the scores that are used to color items in the main panel.

Labels.

o Size variation. The higher the weight of an item, the larger the items label
in the main panel. The Size variation slider determines the strength of this
effect.

o Circles and Frames. These radio buttons determine how items are
indicated in the main panel. If the Circles radio button is selected, items
are indicated both by their label and by a circle. If the Frames radio button
is selected, items are indicated by their label displayed within a rectangular
frame.

o Max. length. This text box determines the maximum length of a label
displayed in the main panel. If the length of a label exceeds the maximum
length, only the first part of the label is displayed.

o Font. This drop down list determines the font that is used to display the
labels of items in the main panel.

Lines.

9
 o Size variation. The stronger the link between two items, the thicker the
line that is used to display the link in the main panel. The Size variation
slider determines the strength of this effect.

o Lines. This text box determines the number of links to be displayed in the
main panel using lines. If the number of links in the network exceeds the
number of links to be displayed, only the strongest links are displayed.

o Colored lines. This check box determines whether links are displayed using
gray lines or colored lines.

o Curved lines. This check box determines whether links are displayed using
straight lines or curved lines.

Colors.

o Cluster Colors. This button is available only if the network visualization is

selected. The button offers four options:

Edit. This option is the default choice. Choose this option to edit the
current cluster colors in the Edit Cluster Colors dialog box.

Import. Choose this option to import cluster colors from a cluster colors
file (see Section 3.4).

Export. Choose this option to export the current cluster colors to a

cluster colors file (see Section 3.4).

Restore original. Choose this option to restore the original cluster

colors.

o Overlay Colors. This button is available only if the overlay visualization is

selected. The button offers four options:

Set range. This option is the default choice. Choose this option to bring
up the Set Overlay Colors Range dialog box. This dialog box can be
used to change the minimum and maximum score that determine how
scores are mapped to colors. By default, scores less than or equal to the
minimum score are mapped to blue, scores equal to the average of the
minimum and the maximum score are mapped to green, and scores
greater than or equal to the maximum score are mapped to red. The
Set Overlay Colors Range dialog box can also be used to determine
how scores are normalized.

Import. Choose this option to import overlay colors from an overlay

colors file (see Section 3.4).

10
 Export. Choose this option to export the current overlay colors to an
overlay colors file (see Section 3.4).

Restore original. Choose this option to restore the original overlay

colors.

o Black background. This check box determines whether the main panel has
a white or a black background color.

When the density visualization is selected, the following options may be available:

Visualization.

o Scale. This slider determines the size of the labels of items in the main
panel.

o Weights. When items have multiple weights, the Weights drop down list
determines the weights that are used in the calculation of the density of
items.

Labels. These options are identical to the options provided when the network
visualization or the overlay visualization is selected (except that the Circles
and Frames radio buttons are not available).

Density.

o Kernel width. This slider determines the value of the kernel width
parameter. We refer to Van Eck and Waltman (2010) for more
information about this parameter.

o Item density or Cluster density. These radio buttons determine

whether the item density visualization or the cluster density visualization
is selected.

Colors.

o Density Colors. This button is available only if the item density

visualization is selected. The button offers three options:

Import. This option is the default choice. Choose this option to import
density colors from a density colors file (see Section 3.4).

Export. Choose this option to export the current density colors to a

density colors file (see Section 3.4).

Restore original. Choose this option to restore the original density

colors.

11
 o Cluster Colors. This button is available only if the cluster density
visualization is selected. The button offers four options:

Edit. This option is the default choice. Choose this option to edit the
current cluster colors in the Edit Cluster Colors dialog box.

Import. Choose this option to import cluster colors from a cluster colors
file (see Section 3.4).

Export. Choose this option to export the current cluster colors to a

cluster colors file (see Section 3.4).

Restore original. Choose this option to restore the original cluster

colors.

o White background. This check box is available only if the cluster density
visualization is selected. The check box determines whether the main panel
has a black or a white background color.

2.3 Information panel

The information panel is available only if item descriptions are available. Item
descriptions may for example have been made available using the description
column in a map file (see Section 3.1). In the information panel, the description of
an item in the currently active map can be shown. For example, when the mouse
pointer is moved over the label of an item in the main panel, the description of the
item will be shown in the information panel.

2.4 Overview panel

In the overview panel, an overview of the currently active map is presented. Each
item in the map is indicated by a small colored dot. A rectangular frame is
displayed in the overview panel to indicate the area in the currently active map
that is shown in the main panel. By left-clicking in the overview panel, the area in
the currently active map that is shown in the main panel can be changed.

2.5 Action panel

The action panel can be used to undertake all kinds of actions. The panel consists
of three tabs: The File tab, the Items tab, and the Analysis tab. These tabs are
discussed in Subsections 2.5.1, 2.5.2, and 2.5.3.

12
2.5.1 File tab

The File tab can be used to perform a number of basic actions. The following
buttons are available on the File tab:

Map.

o Create. Use this button to create a new map. The button brings up the
Create Map wizard. There are three ways in which a new map can be
created using this wizard:

Create a map based on network data. This option requires a network.

The network indicates which pairs of items are linked to each other, and
for each pair of linked items it indicates the strength of their link. To
create a map based on network data, a network file and optionally also a
map file are needed. For a detailed discussion of these files, we refer to
Sections 3.1 and 3.2. Instead of map files and network files, it is also
possible to work with GML files and Pajek files.2

Create a map based on bibliographic data. This option requires

bibliographic data. The data can be read from Web of Science, Scopus,
PubMed, or RIS files. Using this option, it is possible to create maps of
scientific publications, scientific journals, researchers, research
organizations, countries, or keywords based on co-authorship links (i.e.,
multiple items co-authoring the same publication), co-occurrence links
(i.e., multiple items co-occurring in the same publication), citation links
(i.e., one item citing another item), bibliographic coupling links (i.e.,
multiple items citing the same publication), or co-citation links (i.e.,
multiple items being cited by the same publication).3

Create a map based on text data.4 This option requires a text corpus.
The corpus can be read from a corpus file. A corpus file is a text file that
contains on each line the text of a document. This text is assumed to be

2
GML (graph modeling language) files are supported by various computer programs for network
analysis and visualization, for example by Gephi, a popular computer program for network visualization
available at https://gephi.org. Pajek is a well-known computer program for social network analysis (De
Nooy, Mrvar, & Batagelj, 2011). It is available at http://pajek.imfm.si/doku.php. VOSviewer supports
Pajek matrix, network, partition, and vector files.

3
When a map is created based on bibliographic data, either the full counting method or the fractional
counting method needs to be used. We refer to Perianes-Rodriguez, Waltman, and Van Eck (2016) for
more information about these two counting methods.

4
We refer to Van Eck and Waltman (2011) for more information about this option.

13
 in English. Using natural language processing techniques, VOSviewer
extracts terms from the corpus file, where a term is defined as a
sequence of nouns and adjectives (ending with a noun). Based on the
extracted terms, VOSviewer creates a term map. This is a map in which
terms are located in such a way that the distance between two terms
provides an indication of the number of co-occurrences of the terms. In
general, the smaller the distance between two terms, the larger the
number of co-occurrences of the terms. Two terms are said to co-occur
if they both occur on the same line in the corpus file.

In addition to a corpus file, a scores file may also be provided. A scores

file is a text file that contains on each line the score of a document.
Based on the scores in a scores file, VOSviewer calculates a score for
each term in the term map. The score of a term equals the average
score of the documents in which the term occurs. In the overlay
visualization, colors are used to indicate the scores of terms.

It is also possible to create a term map based on Web of Science,

Scopus, PubMed, or RIS files instead of a corpus file. In that case, terms
are extracted from the titles and abstracts of scientific publications.

o Open. Use this button to open an existing map. The button brings up the
Open Map dialog box. To open a map, a map file and optionally also a
network file are needed. For a detailed discussion of these files, we refer to
Sections 3.1 and 3.2. Instead of map files and network files, it is also
possible to work with GML files and Pajek files.

o Save. Use this button to save the currently active map. The button brings
up the Save Map dialog box. A map can be saved in a map file and a
network file. For a detailed discussion of these files, we refer to Sections 3.1
and 3.2. Instead of map files and network files, it is also possible to work
with GML files and Pajek files.

o Print. Use this button to print a screenshot of the main panel.

o Screenshot. This button offers three options:

Save to file. This option is the default choice. Choose this option to
save a screenshot of the main panel. The screenshot resembles the
main panel as closely as possible. However, the navigation buttons in
the upper right corner of the main panel are not shown in the
screenshot. If the Optimize labeling check box in the Screenshot
Options dialog box (see below) is checked, the number of labels visible

14
 in the screenshot is optimized. This means that some labels not visible
in the main panel may be visible in the screenshot. Screenshots can be
saved in a number of graphic file formats. For most applications, we
recommend the PNG format. Some formats, such as EPS, PDF, and SVG,
use vector graphics when saving a screenshot. This has the advantage
that the screenshot can be resized without loss of quality.

Copy to clipboard. Choose this option to copy a screenshot of the main

panel to the clipboard. The screenshot can for example be pasted into a
Word document or a PowerPoint presentation.

Screenshot options. Choose this option to bring up the Screenshot

Options dialog box. This dialog box can be used to change some
screenshot-related settings. The Scaling drop down list determines the
resolution (i.e., the number of pixels) of a screenshot. The resolution is
calculated relative to the resolution of the main panel. Using a scaling of
100%, screenshots have the same resolution as the main panel. Using
the default scaling of 200%, screenshots have a resolution that is twice
as high (i.e., twice as many pixels horizontally and vertically) as the
resolution of the main panel. The Scaling drop down list has no effect
on screenshots that are saved in a file format that uses vector graphics.
The Optimize labeling check box determines whether the number of
labels visible in a screenshot is optimized. Optimizing the number of
labels visible in a screenshot means that some labels not visible in the
main panel itself may be visible in a screenshot of the main panel. The
Include border check box determines whether a border is included
around a screenshot.

Info.

o Manual. Use this button to open the VOSviewer manual. This requires an
internet connection.

o About VOSviewer. This button brings up the About VOSviewer dialog

box. This dialog box provides some general information about VOSviewer,
such as the version number, a copyright notice, a license text, the address
of the VOSviewer web site, and a list of software libraries used by
VOSviewer.

2.5.2 Items tab

The Items tab provides a list of items in the currently active map. By default, a list
of all items in the map is provided. However, a filter can be used to restrict the list

15
to a subset of the items in the map. To do so, enter a filter string in the Filter text
box. This yields a list of all items whose label contains the filter string.

The Group items by cluster check box determines how items are listed. If the
check box is unchecked, items are simply listed alphabetically. If the check box is
checked, items are first grouped by cluster and then listed alphabetically within
each cluster.

By double-clicking on an item on the Items tab, the item is shown in the main
panel.

2.5.3 Analysis tab

The Analysis tab can be used to update the layout and clustering of the currently
active map. This is done using the VOS layout technique and the VOS clustering
technique.5 The Analysis tab can also be used to change the parameters of these
techniques. The following parameters and controls are available on the Analysis
tab:

Normalization. Use the normalization method drop down list to determine

how the strength of the links between items is normalized. The following
options are available:

o No normalization. If this option is selected, no normalization is performed.

We generally do not recommend this option.

o Association strength. If this option is selected, the association strength

method is used for normalizing the strength of the links between items.
Apart from a multiplicative constant, this method is identical to Eq. (6) in
Van Eck and Waltman (2009). This option is selected by default.

o Fractionalization. If this option is selected, the fractionalization method is

used for normalizing the strength of the links between items. Apart from a
multiplicative constant, this method is identical to Eq. (13) in Van Eck and
Waltman (2009).

o LinLog/modularity. If this option is selected, normalization is performed

in the same way as in the LinLog layout technique and the modularity
clustering technique. For more information about these techniques, we refer
to Newman (2004) and Noack (2007, 2009).

5
Together, these two techniques provide a unified framework for layout and clustering. We refer to Van
Eck, Waltman, Dekker, and Van den Berg (2010), Waltman, Van Eck, and Noyons (2010), Waltman and
Van Eck (2013), and Van Eck and Waltman (2014) for more information about these techniques.

16
 Layout.

o Attraction and Repulsion. These parameters influence the way in which

items are located in a map by the VOS layout technique. The Attraction
parameter must have an integer value between -9 and +10. The Repulsion
parameter must have an integer value between -10 and +9. The value of
the Repulsion parameter must be lower than the value of the Attraction
parameter. For most purposes, our recommendation is to set the
Attraction and Repulsion parameters to values of, respectively, 2 and 1.
Values of 2 and 0 or values of 1 and 0 also sometimes yield good results.

o Use default values. This check box determines whether default values of
the attraction and repulsion parameters will be used when creating a new
map using the Create Map wizard (see Subsection 2.5.1). If the check box
is checked, default parameters values will be used. These default
6
parameters values depend on the type of map that is created. If the check
box is not checked, the parameter values specified in the Attraction and
Repulsion text boxes will be used.

o Advanced Parameters. This button brings up the Advanced Layout

Parameters dialog box. This dialog box can be used to change a number of
more advanced parameters of the VOS layout technique. The following
parameters are available:

Random starts. This parameter determines the number of times the

optimization algorithm of the VOS layout technique is run. Each time the
optimization algorithm is run, a different layout may be obtained. The
best layout obtained in all runs of the optimization algorithm will be
used as the final layout. The larger the value of the Random starts
parameter, the higher the quality of the final layout that will be obtained.

Max. iterations. This parameter determines the maximum number of

iterations performed by the optimization algorithm of the VOS layout
technique. The larger the value of the parameter, the higher the quality
of the layout that will be obtained. In general, the default value of the
parameter works well and does not need to be changed.

6
The repulsion parameter has a default value of -1 when creating a co-authorship map of researchers
and a default value of 0 when creating a co-occurrence map of keywords or a citation map of
publications. In all other cases, the repulsion parameter has a default value of 1. The attraction
parameter always has a default value of 2.

17
 Initial step size, Step size reduction, and Step size convergence.
These are technical parameters of the optimization algorithm of the VOS
layout technique. The parameters must have values between 0.000001
and 1. The smaller the value of the Step size convergence parameter,
the higher the quality of the layout that will be obtained. In general, the
default values of the parameters work well and do not need to be
changed.

Random seed. This parameter determines the seed of the random

number generator used by the optimization algorithm of the VOS layout
technique. The seed must be a non-negative integer.

o Update Layout. This button is available only if a network is available. Use

this button to update the layout of the currently active map using the VOS
layout technique.

Clustering.

o Resolution. This parameter determines the level of detail of the clustering

produced by the VOS clustering technique. The parameter must have a non-
negative value. The larger the value of the parameter, the larger the
number of clusters that will be obtained. Our recommendation is to try out
different values for the parameter and to use the value that yields the level
of detail considered most satisfactory for the particular application at hand.

o Min. cluster size. This parameter determines the minimum size of clusters
produced by the VOS clustering technique. In principle, each cluster
produced by the VOS clustering technique must include at least the
minimum number of items specified by this parameter. The Min. cluster
size parameter can be useful to simplify the clustering results obtained
from the VOS clustering technique by getting rid of small and uninteresting
clusters.

o Merge small clusters. This check box determines how the VOS clustering
technique handles small clusters. Small clusters are clusters that include
fewer items than the Min. cluster size parameter. If the check box is
checked, small clusters are merged into larger clusters. If the check box is
not checked, small clusters are discarded and their items have no cluster
assignment.

o Advanced Parameters. This button brings up the Advanced Clustering

Parameters dialog box. This dialog box can be used to change a number of

18
 more advanced parameters of the VOS clustering technique. The following
parameters are available:

Random starts. This parameter determines the number of times the

optimization algorithm of the VOS clustering technique is run. Each time
the optimization algorithm is run, a different clustering may be obtained.
The best clustering obtained in all runs of the optimization algorithm will
be used as the final clustering. The larger the value of the Random
starts parameter, the higher the quality of the final clustering that will
be obtained.

Iterations. This parameter determines the number of iterations

performed by the optimization algorithm of the VOS clustering technique.
The larger the value of the parameter, the higher the quality of the
clustering that will be obtained. In general, the default value of the
parameter works well and does not need to be changed.

Random seed. This parameter determines the seed of the random

number generator used by the optimization algorithm of the VOS
clustering technique. The seed must be a non-negative integer.

o Update Clustering. This button is available only if a network is available.

Use this button to update the clustering of the currently active map using
the VOS clustering technique.

Rotate/flip.

o Rotate. Use this button to rotate the currently active map. The Degrees to
rotate parameter determines the number of degrees by which the map is
rotated.

o Flip Horizontally. Use this button to flip the currently active map in
horizontal direction.

o Flip Vertically. Use this button to flip the currently active map in vertical
direction.

19
3 File types
The two primary file types used by VOSviewer are the map file and the network file.
Map files and network files are simple text files that can easily be viewed and
edited using a text editor (e.g., Notepad) or a spreadsheet program (e.g., Excel).
Map files and network files consist of multiple columns. Hence, each line in a map
file or a network file contains multiple information elements. Different information
elements on the same line are separated from each other by a comma, a
semicolon, or a tab. If an information element (e.g., the label of an item) itself
contains a comma or a semicolon, the information element needs to be enclosed
within double quotes.

We discuss map files and network files in more detail in Sections 3.1 and 3.2. Four
additional file types, the thesaurus file, the cluster colors file, the overlay colors file,
and the density colors file, are discussed in Sections 3.3 and 3.4. We note that all
example files mentioned in this chapter are distributed together with VOSviewer.

3.1 Map file

A map file is a text file that contains information about the items in a map. Each
line in a map file corresponds with an item. The only exception is the first line. This
is a header line that indicates what is contained in each of the columns of a map
file. Below, we list the columns that can be used in a map file. For each column, we
provide the column header and we specify what the column contains.

id The ID of an item. Items need to have an ID only if a map file is used in

combination with a network file.

label The label of an item.

sublabel The sublabel of an item. In the main panel, the sublabel of an item is displayed
below the items ordinary label. Sublabels are displayed in a smaller font.

description The description of an item. The description of an item is used to provide

information about the item in the information panel. HTML formatting can be used
in this column.

url The URL of an item. This column can be used to associate a web page with an
item. Clicking on the label of an item in the main panel will cause the web page
associated with the item to be opened in a web browser.

x The horizontal coordinate of an item.

y The vertical coordinate of an item.

cluster The number of the cluster to which an item belongs. Only integers between 1 and
1000 are allowed in this column.

20
weight The weight of an item. Only non-negative numbers are allowed in this column.
The higher the weight of an item, the more prominent the item is presented in the
main panel.

normalized weight The normalized weight of an item. Only non-negative numbers are allowed in this
column. The higher the normalized weight of an item, the more prominent the
item is presented in the main panel. The default presentation of an item is
obtained by setting the items normalized weight to 1.

score The score of an item. In the overlay visualization in the main panel, items can be
colored based on their score, with by default colors ranging from blue to green to
red.

red The red component of the user-defined color of an item. Only integers between 0
and 255 are allowed in this column. In the overlay visualization in the main panel,
items can be colored based on their user-defined color.

green The green component of the user-defined color of an item. Only integers between
0 and 255 are allowed in this column. In the overlay visualization in the main
panel, items can be colored based on their user-defined color.

blue The blue component of the user-defined color of an item. Only integers between 0
and 255 are allowed in this column. In the overlay visualization in the main panel,
items can be colored based on their user-defined color.

In a map file, one always uses only a subset of the above columns. The order in
which the columns are used is not important.

There are a number of restrictions on the columns that are used in a map file:

There must be an id column or a label column. (If there is no label column, the
ID of an item is used as the items label.)

If there is a sublabel column, there must be a label column as well.

The x and y columns must be used together.

The weight column and the normalized weight column cannot be used together.

The score column and the red, green, and blue columns cannot be used
together.

The red, green, and blue columns must be used together.

Furthermore, it is possible to have multiple weight and score columns. Each weight
or score column then must have its own label. For example, one could have a
weight<Links> column, a weight<Documents> column, and a weight<Citations>
column.

For an example of a map file, see the file journal_map.txt.

21
3.2 Network file
A network file is a text file that contains the adjacency matrix of a network. The
adjacency matrix of a network is a square matrix that indicates for each pair of
items in the network the strength of the link between the items. The strength of a
link is given by a non-negative number. If there is no link between two items, the
strength of the link between the items equals zero. VOSviewer requires networks
to be undirected, and it therefore requires adjacency matrices to be symmetrical. If
a network file contains an asymmetrical adjacency matrix, VOSviewer will make
the matrix symmetrical by averaging corresponding elements on both sides of the
main diagonal.

A network file has either a full format or a sparse format:

Full format. The entire adjacency matrix, including elements that are equal to
zero, is stored in the network file. The file consists of n lines and n + 1 columns,
where n denotes the number of items in the network. The element in the ith
row and the jth column of the adjacency matrix is stored on the ith line and in
the (j + 1)th column of the network file. The first column of the network file
contains IDs of items. This column indicates for each row and column of the
adjacency matrix the ID of the corresponding item. For an example of a
network file in full format, see the file journal_network_full.txt.

Sparse format. Only the non-zero elements of the adjacency matrix are stored
in the network file. The file consists of two or three columns. The first two
columns contain IDs of items. The third column contains the non-zero elements
of the adjacency matrix. This column indicates the strength of the link between
the items referred to in the first two columns. If there is no third column, the
strength of the link between the items referred to in the first two columns
always equals one. In the case of a symmetrical adjacency matrix, it is
sufficient to store only half of the matrix (either the upper triangular part or the
lower triangular part) in the network file. For an example of a network file in
sparse format, see the file journal_network_sparse.txt.

A network file is usually used in combination with a map file. For each item ID in
the network file, there must then be a corresponding item ID in the map file.

3.3 Thesaurus file

A thesaurus file is a text file that can be used in the following situations:

When creating a map based on bibliographic data (see Subsection 2.5.1), a

thesaurus file can be used to merge different variants of a source title, an

22
 author name, an organization name, a country name, or a cited reference. This
may for example be useful when the name of a researcher is written in different
ways in different publications (e.g., with first initial only and with all initials). A
thesaurus file can then be used to indicate that different names in fact refer to
the same researcher.

When creating a map based on text data (see Subsection 2.5.1), a thesaurus
file can be used to merge synonyms into a single term. This may be useful not
only for merging different terms referring to the same concept, but also for
merging different spellings of the same term (e.g., behavior and behaviour). It
may also be useful for merging an abbreviation of a term with the term itself.

Each line in a thesaurus file contains a label and indicates an alternative label that
replaces the original label. The only exception is the first line. This is a header line
that indicates what is contained in each of the columns of a thesaurus file. A
thesaurus file must have two columns: A label column and a replace by column.
The label column contains a label. Depending on the situation (see above), a label
may represent a source title, an author name, an organization name, a country
name, a cited reference, or a term. The replace by column contains an alternative
label that replaces the original label. The replace by column may also be empty. In
that case, the original label is not replaced by an alternative one, but instead the
original label is simply ignored. When creating a map based on a corpus file, this
allows a thesaurus file to be used as a kind of stop word list. A thesaurus file may
for example indicate that certain uninteresting terms (e.g., method, result, and
conclusion in the case of a corpus of abstracts of scientific publications) should be
ignored. For examples of thesaurus files, see the files thesaurus_authors.txt and
thesaurus_terms.txt.

3.4 Cluster colors file, overlay colors file, and density colors
file
A cluster colors file is a text file that contains the colors of clusters. Each line in a
cluster colors file corresponds with a cluster. The only exception is the first line.
This is a header line that indicates what is contained in each of the columns of a
cluster colors file. A cluster colors file must have four columns: A cluster column, a
red column, a green column, and a blue column. The cluster column contains
cluster numbers. Only integers between 1 and 1000 are allowed in this column.
The red, green, and blue columns contain the red, green, and blue components of
the colors of clusters. Only integers between 0 and 255 are allowed in these
columns. For an example of a cluster colors file, see the file cluster_colors.txt.

23
An overlay colors file is similar to a cluster colors file except that instead of a
cluster column it has a color value column, with values between 0 and 1. In the
overlay visualization in the main panel, items can be colored by transforming their
scores into color values and by matching these color values with the color values in
the color value column of the overlay colors file. Exact matching of color values is
usually not possible, and in that case the colors in the overlay colors file are
interpolated. For an example of an overlay colors file, see the file
overlay_colors.txt.

A density colors file is identical to an overlay colors file. In the item density
visualization in the main panel, the color of a point in a map is determined by
transforming the density of items at that point into a color value and by matching
this color value with the color values in the color value column of the density colors
file. Exact matching of color values is usually not possible, and in that case the
colors in the density colors file are interpolated. For an example of a density colors
file, see the file density_colors.txt.

24
4 Advanced topics
In this chapter, some advanced topics are addressed. We first consider the use of
command line parameters (Section 4.1). We then discuss how a map can be made
available online (Section 4.2) and how the amount of memory available to
VOSviewer can be increased (Section 4.3).

4.1 Using command line parameters

VOSviewer supports a large number of command line parameters. These
parameters can for example be used to automatically open a map when VOSviewer
is started or to override some of the default settings of VOSviewer. The command
line parameters supported by VOSviewer are listed below.

Command line parameters for opening or creating a map

gml Use this parameter to specify a GML file. Based on this file, a map will be
opened or created when VOSviewer is started.

map Use this parameter to specify a map file (see Section 3.1). Based on this file, a
map will be opened or created when VOSviewer is started.

network Use this parameter to specify a network file (see Section 3.2). A network will
be read from this file when VOSviewer is started.

pajek_network Use this parameter to specify a Pajek network (or matrix) file. Based on this
file, a map will be opened or created when VOSviewer is started.

pajek_partition Use this parameter to specify a Pajek partition file. Cluster numbers of items
will be read from this file when VOSviewer is started.

pajek_vector Use this parameter to specify a Pajek vector file. Weights of items will be read
from this file when VOSviewer is started.

Command line parameters for creating a term map based on a corpus file

corpus Use this parameter to specify a corpus file. Based on this file, a term map will
be created when VOSviewer is started.

counting_method Use this parameter to specify the counting method to be used when a term
map is created when VOSviewer is started (1 for binary counting and 2 for full
counting).

min_n_occurrences Use this parameter to specify the minimum number of occurrences a term must
have to be included in a term map.

n_terms Use this parameter to specify the number of terms to be included in a term
map. VOSviewer will select the terms that seem most relevant.

scores Use this parameter to specify a scores file. When a term map is created when
VOSviewer is started, this file will be used to calculate a score for each term.

25
thesaurus Use this parameter to specify a thesaurus file (see Section 3.3). When a term
map is created when VOSviewer is started, this file will be used to merge
synonyms into a single term.

Command line parameters for saving a map

save_gml Use this parameter to specify a GML file. When a map is opened or created
when VOSviewer is started, the map will be saved in this file.

save_map Use this parameter to specify a map file (see Section 3.1). When a map is
opened or created when VOSviewer is started, the map will be saved in this
file.

save_network Use this parameter to specify a network file (see Section 3.2). When a map is
opened or created when VOSviewer is started, the network will be saved in this
file.

save_pajek_network Use this parameter to specify a Pajek network (or matrix) file. When a map is
opened or created when VOSviewer is started, the map will be saved in this
file.

save_pajek_partition Use this parameter to specify a Pajek partition file. When a map is opened or
created when VOSviewer is started, the cluster numbers of the items in the
map will be saved in this file.

save_pajek_vector Use this parameter to specify a Pajek vector file. When a map is opened or
created when VOSviewer is started, the weights of the items in the map will be
saved in this file.

Command line parameters for saving a screenshot

save_screenshot_bmp Use this parameter to specify a BMP file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_emf Use this parameter to specify an EMF file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_eps Use this parameter to specify an EPS file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_gif Use this parameter to specify a GIF file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_jpg Use this parameter to specify a JPG file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_pdf Use this parameter to specify a PDF file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_png Use this parameter to specify a PNG file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_svg Use this parameter to specify an SVG file. When a map is opened or created
when VOSviewer is started, a screenshot of the map will be saved in this file.

save_screenshot_swf Use this parameter to specify an SWF file. When a map is opened or created

26
 when VOSviewer is started, a screenshot of the map will be saved in this file.

Command line parameters related to the layout and clustering techniques

attraction Use this parameter to specify the initial value of the Attraction text box on the
Analysis tab in the action panel.

largest_component Use this parameter to indicate that only the largest component of the network
must be kept when running the layout technique when VOSviewer is started.

merge_small_clusters Use this parameter to specify the initial value of the Merge small clusters
check box on the Analysis tab in the action panel (true for checked and false
for unchecked).

min_cluster_size Use this parameter to specify the initial value of the Min. cluster size text box
on the Analysis tab in the action panel.

repulsion Use this parameter to specify the initial value of the Repulsion text box on the
Analysis tab in the action panel.

resolution Use this parameter to specify the initial value of the Resolution text box on
the Analysis tab in the action panel.

run_clustering Use this parameter to indicate that the clustering technique must be run when
a map is opened or created when VOSviewer is started (even if a clustering is
already available).

run_layout Use this parameter to indicate that the layout technique must be run when a
map is opened or created when VOSviewer is started (even if a layout is
already available).

skip_clustering Use this parameter to indicate that the clustering technique must not be run
when a map is opened or created when VOSviewer is started.

Visualization-related command line parameters

black_background Use this parameter to specify the initial value of the Black background check
box in the options panel (true for checked and false for unchecked).

circles_frames Use this parameter to specify how items will be presented in the network and
overlay visualizations (1 for circles and 2 for frames).

cluster_colors Use this parameter to specify a cluster colors file (see Section 3.4). Cluster
colors will be imported from this file when VOSviewer is started.

colored_lines Use this parameter to specify the initial value of the Colored lines check box
in the options panel (true for checked and false for unchecked).

curved_lines Use this parameter to specify the initial value of the Curved lines check box in
the options panel (true for checked and false for unchecked).

density Use this parameter to specify whether the item density visualization or the
cluster density visualization will be selected in the density visualization (1 for
item density visualization and 2 for cluster density visualization).

density_colors Use this parameter to specify a density colors file (see Section 3.4). Density
colors will be imported from this file when VOSviewer is started.

27
density_visualization Use this parameter to select the density visualization when VOSviewer is
started.

kernel_width Use this parameter to specify the initial value of the Kernel width slider in the
options panel.

label_size_variation Use this parameter to specify the initial value of the Size variation slider for
labels in the options panel.

line_size_variation Use this parameter to specify the initial value of the Size variation slider for
lines in the options panel.

max_label_length Use this parameter to specify the initial value of the Max. length text box in
the options panel.

max_score Use this parameter to specify the initial value of the Max. score text box in the
Set Overlay Colors Range dialog box.

min_score Use this parameter to specify the initial value of the Min. score text box in the
Set Overlay Colors Range dialog box.

n_lines Use this parameter to specify the initial value of the Lines text box in the
options panel.

network_visualization Use this parameter to select the network visualization when VOSviewer is
started.

overlay_colors Use this parameter to specify an overlay colors file (see Section 3.4). Overlay
colors will be imported from this file when VOSviewer is started.

overlay_visualization Use this parameter to select the overlay visualization when VOSviewer is
started.

scale Use this parameter to specify the initial value of the Scale slider in the options
panel.

scores_normalization Use this parameter to specify how scores are normalized (1 for no
normalization, 2 for normalization by dividing by the mean, 3 for normalization
by subtracting the mean, and 4 for normalization by subtracting the mean and
dividing by the standard deviation).

show_item Use this parameter to zoom in on a specific item when VOSviewer is started.
The item is indicated by its ID.

white_background Use this parameter to specify the initial value of the White background check
box in the options panel (true for checked and false for unchecked).

zoom_level Use this parameter to specify the initial zoom level in the main panel. The
zoom level must have a value of at least 1. The higher the zoom level, the
more the main panel will be zoomed in on the center of a map.

Miscellaneous command line parameters

encoding Use this parameter to specify the character encoding that is used by VOSviewer
to read and write text files. For a list of the available encodings, see
http://docs.oracle.com/javase/6/docs/technotes/guides/intl/encoding.doc.html.
If this parameter is not used, VOSviewer will attempt to automatically

28
 determine the correct encoding when reading a text file (which in some cases
may result in the use of an incorrect encoding), while it will use the default
encoding when writing a text file.

file_location Use this parameter to specify the folder that is used by VOSviewer as the
default file location.

To use the above command line parameters, VOSviewer needs to be run from the
command line. When using the Windows executable of VOSviewer, this can for
example be done as follows:

VOSviewer map map.txt density_visualization zoom_level 2.5

In this way, VOSviewer will be started and the map in the map file map.txt will be
opened. Also, the density visualization will be selected, and the main panel will be
zoomed in on the center of the map. When instead of the Windows executable of
VOSviewer the VOSviewer JAR file is used, VOSviewer can for example be run in
the following way:

java jar VOSviewer.jar map map.txt density_visualization

zoom_level 2.5

We note that some command line parameters cannot be used together. For
example, the map and pajek_network parameters and the map and corpus
parameters cannot be used together. On the other hand, some parameters can
only be used in combination with another parameter. The pajek_partition
parameter for example can only be used in combination with the pajek_network
parameter. Similarly, the counting_method parameter can only be used in
combination with the corpus parameter.

4.2 Making a map available online

Suppose one wants to make a map available online. One way to do this is simply
by making a map file and a network file (or only a map file) available on the
internet. To open the map, one then needs to take two steps. In the first step, the
map file and the network file are downloaded from the internet. In the second step,
VOSviewer is started and the downloaded map file and network file are used to
open the map.

The above two-step approach is not very convenient, and therefore an alternative
approach is available as well. In this alternative approach, a map can be opened in
VOSviewer directly from a web page. The alternative approach works as follows. A
map file and a network file (or only a map file) need to be made available on the
internet. Suppose these files are located at

29
http://www.example.com/map.txt

and

http://www.example.com/network.txt

The map can then be made available on a web page by creating a hyperlink that
points to

http://www.vosviewer.com/vosviewer.php?map=http://www.example.com/map.
txt&network=http://www.example.com/network.txt

Using this hyperlink, the map can be opened directly in VOSviewer.

We note that all command line parameters discussed in Section 4.1 can also be
used in a hyperlink. For example, to open a map, to select the density visualization,
and to zoom in on the center of the map, the following hyperlink can be used:

http://www.vosviewer.com/vosviewer.php?map=http://www.example.com/map.
txt&network=http://www.example.com/network.txt&density_visualizatio
n&zoom_level=2.5

4.3 Increasing the availability of memory

When using VOSviewer to work with large maps (e.g., maps containing several
thousands of items), the memory requirements can be quite substantial.
VOSviewer will produce an out of memory error if there is not enough memory
available. The availability of memory can be increased by running the VOSviewer
JAR file from the command line and specifying the amount of memory that needs
to be available. For example, if 2000 MB of memory needs to be available, the
VOSviewer JAR file can be run as follows:

java Xmx2000m -jar VOSviewer.jar

When working with large maps, VOSviewer may also produce a stack overflow
error. To prevent this error from occurring, the stack size needs to be increased.
This can be done by running the VOSviewer JAR file from the command line in the
following way:

java -Xss1000k -jar VOSviewer.jar

In this case, the stack size is set to 1000 KB, but other values are possible as well.

30
References
De Nooy, W., Mrvar, A., & Batagelj, V. (2011). Exploratory social network analysis
with Pajek (2nd ed.). Cambridge University Press.

Newman, M.E.J. (2004). Fast algorithm for detecting community structure in

networks. Physical Review E, 69, 066133.

Noack, A. (2007). Energy models for graph clustering. Journal of Graph Algorithms
and Applications, 11(2), 453480.

Noack, A. (2009). Modularity clustering is force-directed layout. Physical Review E,

79, 026102.

Perianes-Rodriguez, A., Waltman, L., & Van Eck, N.J. (2016). Constructing
bibliometric networks: A comparison between full and fractional counting.
arXiv:1607.02452.

Van Eck, N.J., & Waltman, L. (2009). How to normalize cooccurrence data? An
analysis of some well-known similarity measures. Journal of the American
Society for Information Science and Technology, 60(8), 16351651.

Van Eck, N.J., & Waltman, L. (2010). Software survey: VOSviewer, a computer
program for bibliometric mapping. Scientometrics, 84(2), 523538.

Van Eck, N.J., & Waltman, L. (2011). Text mining and visualization using
VOSviewer. ISSI Newsletter, 7(3), 5054.

Van Eck, N.J., & Waltman, L. (2014). Visualizing bibliometric networks. In Y. Ding,
R. Rousseau, & D. Wolfram (Eds.), Measuring scholarly impact: Methods and
practice (pp. 285320). Springer.

Van Eck, N.J., Waltman, L., Dekker, R., & Van den Berg, J. (2010). A comparison
of two techniques for bibliometric mapping: Multidimensional scaling and VOS.
Journal of the American Society for Information Science and Technology,
61(12), 24052416.

Waltman, L., & Van Eck, N.J. (2013). A smart local moving algorithm for large-
scale modularity-based community detection. European Physical Journal B,
86(11), 471.

31
Waltman, L., Van Eck, N.J., & Noyons, E.C.M. (2010). A unified approach to
mapping and clustering of bibliometric networks. Journal of Informetrics, 4(4),
629635.

32