5.1. Global Collaboration of Documents
Table 1 shows the productivity rating of the top ten countries based on the quantity of documents from the Web of Science dataset. In this list, the top four countries were from Asia (China, India, Saudi Arabia, and Pakistan). Furthermore, there were four countries (the UK, Spain, Italy, and Poland) from Europe, the USA from North America, and Australia from Oceania. The result shows that the United States was ranked first among the ten leading countries with 377 publications, followed by the second top-ranked country, China with 332, the UK with 127, and India with 108 papers. Remarkably, China (1316) and the USA (821) had the highest number of citations.
A network of 50 out of 106 countries presented a minimum of three documents, whereas a country with at least five citations in three clusters was to be considered; only 47 countries met the set threshold and were visually mapped using the VOS viewer, as shown in
Figure 2. The co-authorship analysis of countries reveals the type and degree of collaboration in this subject as well as the relationship between the countries involved. Different colors are used to represent each cluster. The number of collaborations between the two nations is represented by the thickness of the link, while the size of the node symbolizes the number of publications emanating from that country [
39].
Additionally, this visualization demonstrates the effectiveness of international cooperation. Regarding social media, the United States, China, England, and India are leading the way. In
Figure 2, the red cluster shows the close collaboration between the USA, Australia, England, and other Asian countries (Saudi Arabia, Pakistan, Japan, and Taiwan). The red and green clusters are the ones with the most significant number of countries (27) and are led by the USA and the Netherlands. They are followed by the blue cluster, with 16 countries, including Switzerland, Spain, Italy, Germany, and Belgium, with Switzerland leading. The countries with the highest number of links were the USA with 377 documents and 1177 links, followed by China (413 links) and India (466 links). The USA was in the first position with total link strength of 173 followed by China (132 total link strengths), the UK (124 total link strengths), Australia (88 total link strengths), and Spain (80 total link strengths) (see
Table 1).
5.3. Affiliation Productivity
Table 2 depicts a scientometric profile of the top 10 most productive affiliation institutes in social media publications. Approximately 1286 institutions were identified from the literature. It was found that a maximum of 14 and 13 publications were published by the University of Gondar and National Cheng Kung University, respectively. A minimum of 10 publications were published by the Harvard Medical School, Hong Kong Polytechnic University, and King Abdul-Aziz University. The results show that the top three institutions were in Hong Kong (Monash University, University of Hong, and Hong Kong Polytechnic University).
The closeness of the relationships is also indicated by the lines connecting each affiliation on the map. The stronger the link between the affiliations, the thicker the line [
40]. The affiliation with the most publications, Monash University with the highest citation received, had the largest font on the map (see
Figure 4). The productivity of affiliations and how they work together with other affiliations are depicted in the figure.
There were approximately 1286 organizations found, but we only looked at the ones that had published at least three research papers. The co-authorship affiliation was analyzed via VOS Viewer software. After applying the filter, 106 institutions were identified, and they were split into six clusters, each of which was represented by a distinct hue. The size of the nodes in the diagram corresponds to the amount of documents, and the strength of the edge in the figure corresponds to the level of collaboration.
Figure 4 provides a VOS viewer of visualization of the 50 most co-authorship affiliations, the threshold quantity of papers for affiliation was set at 3, and only 106 affiliations in six clusters matched the requirement. The size of the nodes in the diagram represents the quantity of documents, and the thickness of the edge in the figure represents the collaboration level [
41]. The first cluster in red is the one with the most significant number of universities (18) and the University of Cambridge in the UK has the largest number of links (n-27 links), followed by the University of Kent (n-23) and the University of Warwick (n-21).
It is followed by the second cluster in green with 10 universities, including University College London, the University of Maryland, and Carolina University. The Minnan Normal University is included in the light blue cluster with seven universities. The University of Melbourne has the largest total link strength (n-30) among the six clusters, followed by Monash University (n-26) and the University of Cambridge with 24 TLS. This indicates that the most significant factor affecting collaborative relationships is not always a considerable advantage.
5.7. Co-Occurrence of Author Keywords
Table 6 highlights the co-occurrence of author keywords for 3147 authors and 1258 indexed keywords by authors in the reviewed publications. It can be seen that the author keyword “COVID 19” has 351 occurrences (564 LS), followed by “social media” with 79 occurrences (351 LS), “pandemic” with 63 occurrences (1267 LS), “corona” with 57 occurrences (145 LS), and “mental health” with 33 occurrences (84 TLS).
We were able to determine the most popular terms and the relationships between them with the assistance of keyword analysis, which led us to the major research questions pertinent to the subject under study. The software was used to evaluate each term and calculate the links, overall link strengths, and co-occurrences of the keyword with other keywords [
52].
To comprehend the core intellectual theme addressed by the existing study, a co-occurrence analysis using 3147 authors and 1258 indexed keywords was performed. “COVID 19” and “social media”, among the author keywords shown in
Figure 5, appeared as prominent terms in the most interconnected platform, which can be addressed by looking at the interconnection links between keywords. The inclusion of keywords in this study was restricted to those that appeared at least three times.
Table 6 displays the TLS for the top 20 keywords and index terms. The most frequently occurring keywords are represented by different nodes in
Figure 5. These frequently used keywords are clustered differently in terms of node size, color, and font size and are linked together by lines based on their similarity [
53]. The number of links between nodes reveals the frequency of those keywords occurring together, and the size of a node is inversely associated with the frequency of a term. The distance between two keywords decreases as the number of co-occurrences between them increases.
The lines of thickness indicate how strong the relationship between keywords is in comparison to the others. The strength of the relationship was determined by the fact that the keywords appeared together in the published articles, including the frequency.
Network visualization was also used to show how often the terms occurred together. As displayed in
Figure 5, 1364 total links in the seven clusters were found during the analysis: in cluster 1 (n-10 occurrences), cluster 2 (n-10 occurrences), cluster 3 (n-8 occurrences), cluster 4 (n-7 occurrences), cluster 5 (n-6 occurrences), cluster 6 (n-5 occurrences), and cluster 7 (n-4 occurrences). In the previous studies, each group of keywords or concepts is represented by various colors using the VOS viewing tool, Cruz-Cárdenas et al. found five clusters [
54], 7 clusters by Al-Zaman [
55], and 5 clusters by Shamsi et al. [
56]. Additionally, all cluster’s keywords were investigated to determine the cluster’s distinguishable theme based on the keywords’ corresponding distinguishable topics.
Cluster 1 (red) has “social media” as a major node in the center and combines other keywords such as “health communication,” and “youtube.” In light of the use of social networks during the COVID-19 pandemic, the cluster is, therefore, connected to SM. These could be considered hot topics in social media in higher education. An infodemic is an information overload during a disease outbreak that includes inaccurate or misleading information through physical and social media.
The phrase “COVID 19” is the center node of Cluster 2 (green). It brings together concepts such as “mental health”, “social distancing”, “stress”, “depression”, and “children.” Hence, this cluster regards the social distance between people, which affected more children and aged people during the COVID-19 pandemic.
The third largest cluster in dark blue, “vaccination”, is prominent in this cluster and groups together other keywords such as “communication” and “twitter”, etc. The COVID-19 vaccination is safe and continues to work well in avoiding fatalities, serious sicknesses, and hospitalizations.
In cluster 4 (yellow), “SARS-CoV-2” represented the closing of all educational institutions during the pandemic, through “video” and “information and communication technology” maintained learning system.
In cluster 5 (purple), the word “social media” is prominently represented by a node. We identified that social media platforms were widely used by institutions in developing nations to maintain e-learning and the educational process [
35].
In the sixth cluster in blue, the keyword of “pandemic” was the link strength found among the top 50 author’s keywords. A pandemic is the global spread of a newly discovered illness. By examining the keywords associated with this cluster through the health information, respiratory viral illnesses with the highest chance of becoming a pandemic include those brought on by the novel influenza virus and COVID-19 [
57].
Cluster 7 (orange), the smallest network cluster, with four keywords, including “attitudes”, “fear”, “knowledge”, and “practice” has no central node.
The research uncovered several groups of significant keywords that are related to one another in smaller groupings. Additionally, “COVID 19” is a core term that serves as a positioning system that is more closely related to other keywords such as “social media”, “pandemic”, and “corona.” In addition to “COVID 19”, “social media”, “pandemic”, and “corona” also frequently co-occur. This finding revealed a significant future study topic for SM-related medical research: the use of new technologies in medical services will grow in popularity and garner more scholarly interest.