Here we present the first study of collaboration network analysis in the academic field of environmental engineering. Scientific collaboration is one of the principal strategies for an individual scholar and a group of scholars to improve the productivity and quality of its research. In this work, we employed coauthor network analysis, which is one of the most powerful means of scientific collaboration network analysis. We collected data for our analysis from papers published from 2007 to 2018 in eight journals whose impact factors were within the top 10% in the subject category of “Engineering, Environmental” or “Environmental Sciences” in 2018 Journal Citation Reports issued by Web of Science and whose official website explicitly states their scope to be topics of environmental engineering. Using the full dataset, 20 clusters, whose members (i.e., individuals) collaborate significantly more intensively with those that belong to the same cluster than the rest, were identified as research communities. Analysis of the features of the members (e.g., keywords of their publications, affiliations) allowed the identification of two major drivers of scientific collaboration: research interest and country of one’s affiliation. We developed a new methodology of network analysis at the mesoscopic level, which was useful in understanding the evolution patterns, such as emergence, convergence, and disappearance, of the research communities. The new methodology includes 1) clustering of networks constructed using sub-dataset at different time intervals and 2) comparing the clusters identified in the sub-dataset with the clusters of the full dataset to track the transition of each cluster over time. This methodology allowed systematically stereotyping the subjects of each cluster as “traditional” or “emerging/ converging”; for example, persistent organic pollutants were determined to be the former and photocatalytic oxidation processes the latter. The yearly growth rate of the volume of publications of the clusters did not depend on the stereotype of the clusters: mean value for the “traditional” clusters was not statistically significantly different from that of the “emerging/converging” clusters (p=0.35, Student’s t-test; 1.26±0.14 and 1.21±0.10, respectively). This result suggested that the emergence of a research community of environmental engineering does not necessarily involve extraordinary quantitative growth in the research output in relevant subjects.

Analysis of collaboration network in environmental engineering: Identification of research communities and their evolution

ACS Spring Meeting 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.