© Wydawnictwo UR 2018 „Edukacja – Technika – Informatyka” nr 1/23/2018

ISSN 2080-9069 www.eti.rzeszow.pl

ISSN 2450-9221 online DOI: 10.15584/eti.2018.1.44

EUGENIA SMYRNOVA-TRYBULSKA1, NATALIIA MORZE2,

OLENA KUZMINSKA3

Mapping and Visualization of a Research Network: Case Study

1
Doktor habilitowany, University of Silesia in Katowice, Faculty of Ethnology and Sciences of
Education in Cieszyn, Head of Department of Humanistic Education and Auxiliary Sciences of
Pedagogy, Poland
2
Profesor doktor habilitowany, Borys Grinchenko Kyiv University, Corresponding Member of
National Academy of Pedagogical Sciences of Ukraine, Vice-rector on Informational Technolo-
gies, Ukraine
3
PhD, University of Life and Environmental Sciences of Ukraine, Department of Information and
Distance Technology, Ukraine

Abstract
The article focuses on trends and methods as well as tools used for mapping and visualization
of scientific domains as a research methodology which is attracting more and more interest from
scientific information and science studies professionals; several factors which have contributed to
such developments are also discussed. Science mapping or bibliometric mapping is a spatial repre-
sentation of how disciplines, fields, specialties, and individual documents or authors are related to
one another (Small, 1999). Different approaches have been developed to extract networks using
the selected units of analysis (authors, documents, journals, terms, etc.). Furthermore, the authors
looks at the most popular computer programs used for the processing of bibliographic and biblio-
metric data.. In addition, plans and activities are presented concerning a comprehensive analysis of
IRNet project (www.irnet.us.edu.pl) networking activities and numerous publications, using spe-
cial ICT tools and study methods, first of all Pajek and presenting a several examples.
Keywords: networking, bibliographic and bibliometric data, mapping, visualization, applications

Introduction
Mapping and visualization of scientific domains as a research methodology
is gaining more and more popularity among scientific information and science
studies professionals. At least two factors have contributed to that. The first is
the digitization of resources, scientific literature and other sources of knowledge,
and hence, growth and development of bibliographic databases available online,
such as ISI Web of Knowledge, Scopus, Google Scholar. The other is a non-linear
increase in computing power, use of supercomputers and distributed technology
(Chen et al., 2010; Zhao, Strotmann, 2008; Osińska, 2010). Conception of the
mapping and visualization of scientific domains as a research methodology pre-
sented on the Figure 1.

327
 Growth and
development of
bibliographic A non-linear
Digitalization of databases available increase in
resources, online, such as ISI computing
scientific literature Web of Knowledge, power, use of
and other sources Scopus, Google supercompu-ters
of knowledge Scholar and distributed
technology

Mapping and visualization of

scientific domains as a
research methodology

Figure 1. Mapping and visualization of scientific domains as a research methodology

Mapping and Visualization of IRNet project Networking Activities

Within the framework of the IRNet project (www.irnet.us.edu.pl) research is
being conducted in several WPs which are separate, yet simultaneously connected
through interrelated stages, which roughly address our several research ques-
tions. We created a bibliographic database containing all the published articles,
books from 2014 to date by all IRNet researchers (178 publications).
This database also includes information regarding the number of authors,
authors' affiliation, country, year, research areas, keywords, the source of publi-
cation (e.g., journal, book, etc.) and certain other data. Consequently, this dataset
enables us to assess the main trends in IRNet scientific publications production.
The time frame of the analysis is the last 4 years of IRNet existence, in which we
have been able to trace its knowledge production and dissemination.
Based on the dynamics of international co-authorships, we will be able to
map and trace international collaboration patterns and thus infer IRNet geo-
graphical influence scope, i.e. its international influence (Research Question).
By means of additional research utilizing information available in the Insti-
tute for Scientific Information (ISI), namely in the Science Citation Index (SCI),
we assess the geographical pattern of the citations of IRNet scientific activities,
publications, production. This enables us to evaluate to what extent IRNet pro-
ject scientific production has been cited at the European and world level.

Methodology of research
The main indicators of research effectiveness – bibliometric indicators – is
a powerful information tool to support the development of science. So today, all

328
over the world, they constitute obligatory elements in the reporting of scientific
institutions, universities, research groups and individual scientists.
In this case, we consider an information analyst as a tool for monitoring scien-
tific information flow within the framework of the international project IRNet,
the influence of the publications of its members in the international scientific
information space, as well as predictive and advisory activities to improve repre-
sentativeness of publications in international scientometric databases, the ap-
proximation of the level of scientific publications of individual scientists to
world standards.
The articles became the theoretical basis of research, they show the current
processes in Biblio – and scientometrics. The article’s metadata of participants
of the project became the practical base and they were placed in the scientomet-
ric database BDWeb of Science and other international journals. Practical re-
search methods were prepared using the programs: Bibexel (metadata analyses),
and program Pajek: the graph editing and visualization of the graph structure
of co-authorship – of the method of Kamada-Kawai (Kamada, Kawai, 1988).
For the study of science as a process of scientific communication, which is
relevant to this study, one can use the tools of network analysis. As a tool for
analysis and visualization of large scale networks we can recommend Pajek
(Pajek, 1996). However, for complex types of analysis, one must have compe-
tencies helping one to work with specialized products and to develop pro-
gramming skills.
To carry out a network analysis of participants of the project we will build
a network of co-authorship for research activity and presentation of results in
scientific publications.
Analyses of data for the purpose of network analysis were performed using
Bibexel, starting with obtaining initial data for this program. These data (up to
end of 2016) were assembled using built-in tools of the scientometric database
Web of Science. Data for each participant in the project IRNet are files with the
extension .txt, which were combined using the program Bibexel: in the new
window Type new/file name here, enter name which needs to have the merged
file, select all the files you want to merge, and run Files -> Append all selected
files to another.
With this program, the files were created for visualization in the Pajek pro-
gram: the file edit Analiz.net (Mapping -> Create net-file for Pajek), and the file
number of publications of each participant of the project.vec (Mapping -> Create
vec file). Because not all publications of the project participants belonged to the
metric databases of the Web of Science database, editing of data for network
analysis was performed manually in Notepad.

329
 The procedures in Pajek's main window are organized in accordance with
the types of data objects that they use as input. Partitions, Vectors, Permutations
can be used to store properties of vertices. Objects of Cluster type define subsets
of vertices according to certain characteristics.
For building a graphics editing program Pajek (http://mrvar.fdv.unilj.si/pajek//)
the following objects were used:
1. Networks are the main objects (vertices and lines), and the default exten-
sion is: .net. In this case, the created file is Analiz.net of the 83 vertices – authors’
articles and links between them, reflecting the co-authorship in publications.
2. Vectors – file Analiz.vec, contains 83 entries, where for each vertex
(author) its quantitative characteristic (real number) – the number of publications
is indicated.
3. Partitions – file Analiz.clu, where for each vertex the class is defined to
which it belongs. In this case, there are 10 classes, in line with the countries’
mission project participants:
To build a network visualization of co-authorship in the program Pajek, it is
necessary to download the file Analiz.net in the Networks field. Vectors –
Analiz.vec Partitions – file Analiz.clu, and select Draw -> Network + First Parti-
tion+ First vector.
The result (Figure 2) obtained is the visualization of the connectivity graph
of the authors in the framework of the project.

Figure 2. Example of network’s visualization of co-authorship

Source: own work.

330
 As for building the visualization of networks linking scientists with special
software and settings, project participants’ “labeled” rectangles (diamond) and
the corresponding vertices have the signatures – the names of the scientists, not
the participants – triangles (Triangle); shapes-marks are of different colors de-
pending on the country, which represent the authors, and the size that is deter-
mined by the number of each author’s publications. As can be seen in the Figure
the scientists have different publication activity, some who are not the partici-
pants of the project have a greater number of publications than some of the par-
ticipants. The latter can be considered as a basis for widening the circle of parti-
cipants or an invitation of active scientists to other projects.
The distribution of the contribution of the project participants by the number
of posts is not uniform: the center is represented by more active members. For
each network node the author's name and the number of publications are added
(which also affects the size of the geometric area). The color is also allocated to
the country offices participating in the project.
It should be noted that the visualization, basically, is a brief synthesis of the
results of the analysis that allows us to understand the context of the transition to
the data that underpin it. The credibility presented in the form of a visualization
of the analysis results will be high, if there is a verification of the results of the
analysis, that is, all the visualization elements retrieved can be traced back to the
primary data and methods of their processing.
The results of this visualization can be used to make optimal decisions on
the management of information resources with the aim of improving processes
of scientific communication and the evaluation publications contribution of each
participant. To determine the scientific impact of the project participants in the
development of pedagogical science an additional analysis needs to be conducted.

Conclusion
This article presents trends, methods as well as tools used for the mapping
and visualization of scientific domains as a research methodology which is gain-
ing more and more popularity among scientific information and science studies
professionals; furthermore, several factors contributing to this development are
discussed.
In the market there is an extensive offering of various types of software for
processing of bibliographic and bibliometric data; software for mapping and
visualization as well as interdisciplinary research – open source license as well
as commercial license programs.
The author of the article, a coordinator of the international IRNet project,
describes certain plans, methodologies and activities relating to comprehensive
review of IRNet project networking activities and numerous publications, using
special ICT tools and study methods; results of these activities will be presented
in subsequent publications.

331
References
Chen, Ch., Ibekwe-Sanjuan, F., Hou, J. (2010). The Structure and Dynamics of Co-citation Clusters:
A Multiple-perspective Co-citation Analysis. Journal of the American Society for Information
Science and Technology, 61 (7), 1386–1409.
Kamada, T., Kawai, S., (1988). An Algorithm for Drawing General Undirected Graphs. Informa-
tion Processing Letters, 31, 7–15.
Osińska, V. (2010). Rozwój metod mapowania domen naukowych i potencjał analityczny w nim
zawarty. zin, 2 (96), 41–51. Pobrane z: http://repozytorium.umk.pl/handle/item/463 (1.02.2017).
Osińska, W. (2016). Wizualizacja informacji. Toruń: Wyd. UMK.
Program Pajek. Pobrane z: http://mrvar.fdv.unilj.si/pajek// (1.02.2017).
Small, H. (1999). Visualizing Science by Citation Mapping. Journal of the American Society for
Information Science, 50 (9), 799–813.
Smyrnova-Trybulska, E. (2017). Mapping and Visualization of Scientific Bibliometric Domains
and Research Network Activities. In: M. Chruby (ed.), DLSC2017 Conference Proceedings
(p. 301–308). Brno: University of Defence.
Zhao, D., Strotmann, A. (2008). Information Science during the first Decade of the Web: An
Enriched Author Cocitation Analysis. Journal of the American Society for Information Science
and Technology, 59 (6), 916–937.

Acknowledgement
The research leading to these results has received, within the framework of the IRNet
project, funding from the People Programme (Marie Curie Actions) of the European
Union's Seventh Framework Programme FP7/2007-2013/ under REA grant agreement
No: PIRSES-GA-2013-612536

332