2.3.2. Network Analysis Methods

VOSviewer is a free software tool based on the Java environment, that is suitable for constructing complex networks using large-scale data. Therefore, we used VOSviewer software (version 1.6.9) to conduct an in-depth network analysis to visualize the connections between various MPs-related items, and to explain their network structure.

Countries co-authorship network: We conducted the co-authorship analysis to identify collaboration networks among different countries in the MPs research field. The nodes represented countries contributing to MPs research, and the links between items implied cooperative relationships. The size of the node increased as the number of articles published by an individual country increased. The value of the links indicated the number of times a given country shared co-authorship with others. The strength of the link increased as the number of co-authorships increased.

Cited reference co-citation network: Analyzing a knowledge base in a certain research field can be conducted by co-citation network analysis for the cited references [18]. In our co-citation analysis, the nodes represented scientific references. The node size represented the number of times a reference was cited. The distance between two references indicated the correlation of the articles according to co-citation links, based on the assumption that more frequently co-cited references exhibited greater co-citation strength.

Author keywords co-occurrence network: The keywords that authors provided for their articles about MPs research represented their academic viewpoints. Thus, our author keywords co-occurrence analysis identified important terms in the MPs academic, as well as the research hotspots in the MPs discipline. The nodes represented high-frequency author keywords, and the size of an individual node represented how many times that keyword occurred. The link strength between two nodes indicated the number of articles in which two keywords occurred together.
