K-core analysis and modeling for network centralities

Abstract

Network modeling is the interdisciplinary study of relationships. Network analysis deals with relational data and Network modeling represents the interdisciplinary study of relationships. Network structure can be studied at many different levels. Around 1000 article titles on cancer, published in journal resources were considered as a dataset. Data exploration was done through displaying nodes and edges in various layouts. With a term frequency limit of 100, nearly 64 terms appeared which less than 1% sparse is. Word cloud data was plotted using word frequencies from term matrix data. An undirected network graph plotted and evaluated density, average path length and modularity, which were found to be within limits. K-cores have also been used to analyze the connectivity of a network. Network centralities such as Degree centrality, Closeness, Eigenvector and between’s centrality resulted in node carcinoma_ being more central in the network.