Title: Automatic detection and analysis of long-term changes in travel patterns of public transport passengers using Smart-Card data
Abstract: All around the world, a large number of people rely on public transport. These days, most transport users automatically pay with a smart-card when using public transport systems. Due to people's frequent usage of public transport, researchers are invested in better understanding passengers' behaviour patterns. The smart-card system is an ideal source of data for researchers as it is capable of storing large amounts of data on transit users, and when used for a long time, can be the source needed for detailed investigation into passengers' trip patterns, both habitual and sporadic.There are two types of changes in the variation of passengers' movements. First, there exists a short-term pattern, which is expressed as a momentary variation in movement pattern. Second is more long-term and of specific interest to transport officials who are interested in understanding the underlying factors that may disturb the more regular and longer lasting habits. As such, the nature of this study is designed so as to improve detection and the analysis of temporal long-term changes in passengers' movement patterns by relying on the smart-card system in order to obtain detailed information about passengers, such as their time-of-day and day-of-week behaviour patterns.One of the main problems with using big data, however, is that the processing of it is time-inefficient and expensive when performed manually. The study therefore aims to process the smart-card data using an automated approach where the proposed framework relies on machine learning and pattern recognition techniques. This result strives to aid public transit operators to improve the provision of public transport to commuters by taking better account specific changes in passengers' movement behaviour. The research aims to provide public transit operators information that will serve to improve understanding of commuters' behaviours. As a result, public transit operators may better design customer retention strategies that will improve transit services and provide greater efficiency in inter-regional migration.