The project focuses on Social Network Analysis of transportation systems, using the NetworkX library of Python. More specifically, the analysis of the Indian Eastern Railway network to identify its structural properties, strengths, and weaknesses is the motive of this project. While transportation networks like airplane routes, city traffic routes, and railway networks have been explored in the past, no such study has been performed in recent years on the Indian Railway network in India. For this thesis, data was prepared from scratch using the IRCTC online train timetables. Consolidating the data of nearly 1000 trains and their various stations into one program-readable file is the main motive.

As initial analysis, only trains from and to Howrah, West Bengal, were added to the database, and an empirical analysis of the same was done. Based on these results, more trains were added, and a study is being developed on similar lines to those performed by Ghosh et al. [1], Mohmand and Wang [2], and Cao et al. [3] among others.

An analysis was performed separately on four different datasets having 46 trains, 684 trains, and 967 trains, respectively. The graphs prepared to represent the networks contain the stations as nodes, and the weight of each edge between a set of nodes represents the number of trains connecting those two stations. Various metrics of measurement have been used to study the structure of the developing network, such as centrality measures, the number of triangles, degree distribution, etc.

As the database increased in size and more trains were added to the network, the study of the structure and an analysis of the connectivity between stations and the identification of the most crucial and well-connected stations was done.

Once structural properties were studied, analysis was done to identify crucial stations, possible junctions where more trains are needed, possible sources and destinations which could use alternate routes, as well as finding alternate routes with the shortest distance to be traveled using actual geographical data. To be more specific to users, an algorithm was developed to filter the preferences of a traveler and provide them with routes based on their choices, which could include only a specific number of stoppages as well as avoiding certain stations.