PSI - Issue 53

Carla M. Ferreira et al. / Procedia Structural Integrity 53 (2024) 254–263 Carla M. Ferreira / Structural Integrity Procedia 00 (2019) 000 – 000

258

5

2.3. Data Analysis A total of 396 documents including articles, conference papers, books, book chapters and reviews were considered eligible for bibliometric analysis. To conduct this analysis, the freely available open-source bibliometrix tool was selected to perform a comprehensive science mapping analysis of the results. Bibliometrix uses the R language to perform a bibliometric analysis along with Biblioshiny, a web-based application which combines the functionalities of the bibliometric package with the web interface using the Shiny packaged environment [9]. 2.4. Data Visualization The technique of visualizing bibliometric networks, often referred to as science mapping, has proven to be a valuable and powerful approach to analyze a large variety of data. Several visualization approaches were proposed in the past years for visualizing bibliometric networks, where the three most popular ones are the distance-based approach, the graph-based approach and the time-based approach, among others. In the distance-based approach type, the one of interest for the present work, a bibliometric network is represented by nodes. The nodes are positioned in such way that the distance between two nodes roughly indicates the relatedness between them. In general, the smaller the distance the higher the relatedness and vice-versa. A recent and successful technique for determining the location of the nodes in a distance-based visualization is the VOS technique, developed by Eck et al., and which is used in the VOSViewer software [10]. The VOSViewer software is a tool that offers in a relatively easy way the basic functionalities required for visualizing bibliometric networks. Due to its simplicity in use and reliability, VOSViewer was utilized to visualize the 396 documents previously mention for analysis and the results were interpreted, in accordance with the general science mapping workflow and guidelines previously mentioned. 3. Results 3.1. Descriptive statistics The timespan of the retrieved data comprises documents published between 2003 and July 2023, the data at which the present work was done. These documents were collected from 185 diverse sources and a total of 1537 authors were involved in these publications, in accordance with the bibliometric information depicted in Fig. 2, provided by the Biblioshiny interface, from where it can be concluded that more than half of this dataset are articles.

Description Timespan

Results

2003:2023 (July 2023)

Sources (Journals, Books, etc)

185 396

Documents

Annual Growth Rate %

20.69 1537

Authors

Authors of single-authored docs

6 6

Single-authored docs Co-Authors per Doc

4.69 851

Author ’ s Keywords (DE)

References

13779

Document Average Age Average citations per doc

3.65

31.42

Fig.2. Overview of the final dataset retrieved from Scopus database used for bibliometric analysis.