*3.1. Bibliometric Analysis of SLR Method Results*

Queries 4 and 5 (Table 2) were used for studying the Scopus database with different results for the same time point. The obtained results of those two queries were 671 and 611 publications, respectively (Table 2). Those results were analyzed in the VOSviewer software in form of bibliometric maps representing the keywords frequently occurring together.

Figure 2 is a bibliometric map of keyword co-occurrences of indexed keywords from publications index in Scopus distinguished as the Q4 results (Table 2). The method used to generate Figure 2 was full counting, and in this method 2418 indexed keywords were identified, among them 67 indexed keywords met the threshold of 10 co-occurrences. Among those results, keywords were referring to the countries and organizations' names that were deselected from the proposed keywords list. Additionally deselected keywords from the proposed in VOSviewer list were: "human", "humans", "article", "female", "male", and "adult". Then from 67 keywords, 11 were deselected. Finally, there are 56 keywords collected in four clusters automatically colored and identified by the VOSviewer software. Figure 2 presents the keywords most often used in the scientific publications dedicated to green jobs, green collars, green employment, and sustainable employment and their combinations explored by the Q4 syntax. As a result, only two keywords representing "green jobs" and "green job" were placed in the bibliometric map (Figure 2).

There are four clusters presented in Figure 2 and Table 3, and those clusters were automatically organized by the VOSviewer software. There are different numbers of keywords in each of the four clusters. In the first red-colored cluster there are 23 items and this is the most numerous group of keywords. The second is marked in green in Figure 3; this cluster consists of 15 keywords. The third cluster consists of 13 keywords presented in blue in Figure 2. There is also a yellow cluster with 5 automatically distinguished keywords. At this aggregate level, it is possible to identify themes of clusters of keywords based on the co-occurrence's frequency. The size of nodes presented in Figure 2 is proportional to the number of occurrences of indexed keywords. Another important feature of the presented bibliometric map is the fact that closer proximity between nodes indicates a closer relationship between keywords. These characteristics allow aggregate keywords into clusters presented in Table 3. The number of occurrences for each keyword is indicated in parentheses, after each keyword.

Based on the generated results of Q4 there was also an overlay map generated (Figure 3). The purpose of this figure is to present the evolution of the scientific interests represented by the keywords related to the GJs. Figure 3 has automatically generated a time scale by VOSviewer. Figure 3 is similar to Figure 2 in shape and represents the same nodes, and edges as in Figure 2 and occurrences in Table 3.

In Figure 3 there are visible darker and lighter elements. The dark blue color represents the oldest keywords and this group reflexes the fourth cluster in Table 3. Keywords represented by the yellow nodes in Figure 3 represent the newest and still actual fields of interest in the subject of green jobs, and even the keyword "green job" is still in yellow in Figure 3. The importance of these yellow-marked keywords is the basis of the discussion and conclusions for future research directions in respective sections of this paper.

**Figure 2.** Bibliometric map of keywords co-occurrences Q4 results analysis in Scopus. Source: Authors' elaboration.



Source: Authors' elaboration.

Figure 4 is a bibliometric map of keywords co-occurrences of indexed keywords from publications index in Scopus distinguished as the Q5 results (Table 2). The method used to generate Figure 4 was full counting of indexed keywords co-occurrences, and in this method, 2213 indexed keywords were identified, among them 58 indexed keywords met the threshold of 10 co-occurrences. Among those results, keywords were referring to the countries and organizations' names that were deselected from the proposed keywords list. Additionally deselected keywords from the proposed in VOSviewer list were: "human", "humans", and "article". Then from 58 keywords, 7 were deselected. Finally, there are 51 keywords collected in four clusters, automatically colored and identified by the VOSviewer software. Figure 4 presents the keywords most often used in the scientific publications dedicated to green jobs, green collars, green employment, and sustainable employment and their combinations explored by the Q5 syntax. As result, only two keywords representing "green jobs" and "green job" were placed in the bibliometric map (Figure 4).

**Figure 3.** Overlay Visualization of keywords co-occurrences Q4 results analysis in Scopus. Source: Authors' elaboration.

**Figure 4.** Bibliometric map of keywords co-occurrences Q5 results analysis in Scopus. Source: Authors' elaboration.

There are five clusters presented in Figure 4 and described in Table 4 which are automatically colored by VOSviewer software. There is the most numerous of all clusters colored in red the first cluster with 18 items. Second is the green cluster with 14 distinguished keywords. The third is a blue cluster consists of 9 items and collects the keywords: "recycling", "waste management", "risk assessment", "occupational exposure", "occupational health", occupation", "environmental health", "energy conservation", and "sustainable development". There is also a fourth yellow cluster with 5 items. The fifth cluster in Table 4, consists also of 5 keywords but it is colored purple. The number of occurrences for each keyword is indicated in parentheses, after each keyword, in Table 4.

**Table 4.** Clusters of keywords co-occurrences visible in Figure 4 for Scopus Q5 results.


Source: Authors' elaboration.

There are similarities between the presented two tables with VOSviewer results automatically divided into clusters, although Table 3 consists of more keywords than Table 4. The number of clusters in Table 4 is also smaller than in Table 3. The first cluster in Table 4 revolves around negative aspects of the GJs definition expressed in human activities' pressure on the natural environment measures. The second cluster presented in Table 4 consists of positive aspects of the GJs definition expressed in progress, economic development, and sustainability. There is also a third cluster in Table 4 and this cluster revolves around employee health protection and conservation of the resources. The fourth cluster presented in Table 4 represents the rules or regulations associated with the GJs which influence "energy resources", "environmental health", "labor unions", "organization and management", and "trade union".

There are the same similarities between Figures 4 and 5 as the described similarities between Figures 2 and 3, in terms of the shape and connections. In Figure 5 there is also an automatically distinguished time scale of co-occurring keywords evolution. The oldest keywords marked in darker colors correspond with the fifth subnetwork of the created map and parts of the other clusters. The distribution of those older keywords is then complex. However, the lighter keywords representing the relatively newest scientific interests are scattered. In Figure 5, attention is deserved for two centrally located keywords "green job" and "green jobs" marked in lighter colors, which indicates the ongoing debate which revolves around those terms. Based on Figure 5, the view on the perspective and emerging future directions of studies are developed in the discussion and conclusion sections.

The order of keywords in each cluster is automatically proposed by the VOSviewer software. Their complex relations prove that the concept of GJs emerged from the concept of sustainable development and the assumption that greening the economy or creating a green economy, would generate GJs [61]. Therefore, the most important and biggest node in Figures 2 and 3 is the "employment" keyword. Its central place in the bibliometric map reflects the research results, which claim that with GJs it is possible to fight unemployment as well as to counteract environmental degradation [62].

**Figure 5.** Overlay Visualization of keywords co-occurrences Q5 results analysis in Scopus. Source: Authors' elaboration.

There are not only quantitative differences in Q4 and Q5 results but also qualitative, related to the GJs definition. The broad definition of GJs has raised concerns among the expert teams and created the need to clarify the direction of further work on the definition of GJs. As part of their work on the GJs definition, the teams identified their specific sectors for their regions that meet the condition of respecting nature's assets and residents [63,64]. The colors presented in Figures 2 and 4 are also different, although the keywords in Tables 3 and 4 are similar. The shapes of Figures 2 and 4 are also matching. The most interesting feature of both figures is separated on the right side of the area of the figures which consists of five keywords. These nodes are as follows: "energy resource", "environmental health", "labor unions", "organization and management", and "trade union". These keywords are mainly related to the green labor market or labor conditions and are the same in both bibliometric maps.
