Abstract
Scientific communication via publications in journal articles have been the prime method of engagement and dissemination of work of individual and small groups of scientists throughout history. The rapid growth of scientific sphere and confluence of social and economic factors have led to the number of publications rising dramatically in the last three decades, rendering it impossible for any sole expert to read all papers pertaining to their individual fields. Thus, new methods are required to e.g., understand the direction of scientific research within fields of study, ascertain importance of particular groups, authors, or institutions, compute metrics that can determine the importance (centrality) of particular seminal papers, and provide insight into the social (collaboration) networks that are present. Here, we present such a method based on semantic analysis of citation networks, using the freely available CiteSpace program. We use citation network analysis on three examples, including a single material that has been widely explored in the last decade (BiFeO3), a small sub-fields with minimal number of authors (flexoelectricity and Kitaev physics) and a wide field with thousands of publications pertaining to a single technique (scanning tunneling microscopy). Interpretation of the analysis and key insights on the fields, such as whether the fields are experiencing resurgence or stagnation, are discussed, and author or collaboration networks which are prominent are determined. These insights allow the wealth of literature to be explored to determine the important connections between cited works, particular authors and groups, and the direction of the topic as a whole to be gleamed. Such methods represent a paradigm shift in our way of dealing with the large volume of scientific publications, and can change the way literature searches and reviews are conducted, as well as how impact of specific work is assessed.
