An Empirical Study on the Role of Macro-Meso-Micro Measures in Citation Networks

Abstract

Due to the growing number of articles published every year as the research output, it is imperative to analyze their impact on future and current research. In this cito-analytical study, we use two publicly available citation datasets, i.e., arXiv’s High-Energy Physics Citation Theory Network and Cora Citation Network. This study employs different macro-meso-micro level indicators such as K-cores, centrality measures, and clustering coefficient in identifying relevant network characteristics and establishing their inter-relationships to determine impactful research. While the meso-level feature identifies the type of citation network, the micro-level indicators (centrality measures) help in recognizing the individual node (research paper) strength and macro-level statistics comments upon the global network characteristics. The current exposition empirically demonstrates the relevance of using macro-meso-micro level statistics together as the unit in determining influential and significant research output. While previous researchers have independently used these metrics in other academic networks, we, however, showed their importance and inter-relationship using an integrationist approach in citation networks.