Studying Indian Railways Network using hypergraphs

In Social Network Analysis, most of the systems are modeled using graphs. However, certain problems instances have higher order interactions, i.e., they involve more than two actors in an interaction. When simple graphs are used in these situations, information captured by them lack qualitative information. In this paper we take the example of Indian Railways Network which is one of the biggest railway networks in the world and show how hypergraphs better capture qualitative information than simple graphs while studying the characteristics of the system.We compare the results of hypergraphs and simple graphs in modeling latent underlying ground truth properties such as diameter of the network, community structure, etc. Empirically we show that hypergraphs lead to better models on the Indian railway network.