Emerging properties of the citations network

Abstract

This essay presents an analysis of the American Physical Society (APS) dataset, comprising metadata and citations of APS articles. We focus on network structures behind the dataset. We build three different networks: citation network, author citation networks, and co-author network. We compute key network metrics such as degree distribution, clustering coefficients, and centrality measures. We expect the degree distribution of the different networks to follow a power-law distribution. Through simulations, we compare the citation network structure to randomly generated directed graphs built by the configuration model with the same degree distribution, obtaining substantially different structural configurations. The in-degree distribution of the citation network exhibited scale-free properties, compatible with the preferential attachment network development. This study provides a comprehensive understanding of the APS citation network and emphasizes the importance of temporal and preferential attachment mechanisms in shaping real-world networks