Graph mining for the detection of overcrowding and waste of resources in public transport

Abstract

The imbalance between the quantity of supply and demand in public transport systems causes a series of disruptions in large metropolises. While extremely crowded vehicles are uncomfortable for passengers, virtually empty vehicles generate economic losses for system managers, and this usually comes back to passengers in the form of fare increases. In this article a new data processing methodology will be presented for the evaluation of collective transportation systems. It proposes the construction and mining of graphs that represent complex networks of supply and demand of the system to find such imbalances. In a case study with the bus system of a large Brazilian metropolis, it was shown that the methodology in question is capable of identifying global imbalances in the system based on an evaluation of the weight distributions of the edges of the supply and demand networks. It has also been shown that even in a scenario where information about the demand is incomplete, using community detection techniques it is possible to identify the stretches of the network that are potentially causing these imbalances on a global scale.