Phenomenal Data Mining and Link Analysis

Journal of Transport Geography, 2020

Highlights • Smart card data are used to define public transport passenger profiles. • The study is based on a coastal tourist destination during peak season. • The analysis focuses on travellers using a multipersonal transport fare. • Model-based clustering is used to identify passenger profiles. • The method provides useful information for public transport policy and management. Abstract Data collected through smart travel cards in public transport networks have become a valuable source of information for transport geography studies. During the last two decades, a growing body of literature has used this sort of data source to study the behaviour of public transport users in cities and regions around the world. However, its use has been scarce in contexts where public transport demand is highly influenced by the activities of the tourist sector. Therefore, it remains to be seen whether these data can be leveraged to optimize the supply of public transport. In this article, data drawn from the Camp de Tarragona automated fare collection system extracted during 2018 are used to study tourists' use of public transport in Costa Daurada (Catalonia, Spain). This is a popular coastal destination with a high concentration of visitors during the summer period. The analysis f ocuses on the use of the T-10, a multipersonal transport fare with no time limitations on its use which makes it appealing for tourists. Model-based clustering has been applied to identify different clusters of passengers according to their activity and spatial profiles. Differences between profiles are significant and, as a result, this study allowed the validation of a method that could be replicated in other contexts, as it provides highly useful information for public transport policy and mobility management.

Transportation Research Part C: Emerging Technologies, 2017

In recent years, there has been increased interest in using completely anonymized data from smart card collection systems to better understand the behavioural habits of public transport passengers. Such an understanding can benefit urban transport planners as well as urban modelling by providing simulation models with realistic mobility patterns of transit networks. In particular, the study of temporal activities has elicited substantial interest. In this regard, a number of methods have been developed in the literature for this type of analysis, most using clustering approaches. This paper presents a two-level generative model that applies the Gaussian mixture model to regroup passengers based on their temporal habits in their public transportation usage. The strength of the proposed methodology is that it can model a continuous representation of time instead of having to employ discrete time bins. For each cluster, the approach provides typical temporal patterns that enable easy interpretation. The experiments are performed on five years of data collected by the Société de transport de l'Outaouais. The results demonstrate the efficiency of the proposed approach in identifying a reduced set of passenger clusters linked to their fare types. A five-year longitudinal analysis also shows the relative stability of public transport usage.

2016

Direct and easy access to public transport information is an important factor for improving the satisfaction and experience of transport users. In the future, public transport information systems could be turned into personalized recommender systems which can help riders save time, make more effective decisions and avoid frustrating situations. In this paper, we present a predictive study of the mobility patterns of public transport users to lay the foundation for transport information systems with proactive capabilities. By making use of travel card data from a large population of bus riders, we describe algorithms that can anticipate bus stops accessed by individual riders to generate knowledge about future transport access patterns. To this end, we investigate and compare different prediction algorithms that can incorporate various influential factors on mobility in public transport networks, e.g., travel distance or travel hot spots. In our evaluation, we demonstrate that by com...

Irish Signals and Systems Conference, 2011

A novel methodology to infer transportation mode taken by mobile device users between regions of interest is introduced. It relies on analysing anonymised billing data, namely call detail records, supplied by mobile network operators as the primary source of user-created data. Coupled with the spatial coverage and distribution of mobile network cells and geographical route map information of major transportation modes, assumed to be partially non-overlapping, user travel paths can be predicted. Journey specific trajectories are constructed and analysed using the concept of virtual cell path for each qualified pre-processed list of activities from each unique user. After classification, kernel density paths for each route were generated both for illustration and validation purposes. Differentiation between rail and road users travelling between Dublin and Cork in the Republic of Ireland is shown as an example application case study.

Proceedings of the National Academy of Sciences, 2013

2019

The study of temporal patterns has been applied to represent the various behaviours of transport users. Smart card data is useful for characterizing travel behaviours. The behaviours identified can be analysed and thereby transport services can be improved. For large datasets of transactions, the traditional method is to segment the data into several groups and use one unique pattern to represent each group. However, classifying very large datasets is still challenging. Here we propose a method to classify the temporal patterns of all the users of a public transit system. We recommend a clustering method that combines a sampling method and cross-correlation distances. This method was applied to classify the temporal patterns of public transport users from Gatineau, Canada. An indicator was developed to validate the efficiency of the proposed method. Compared to current methods, the proposed method is faster and better able to deal with very large datasets.

17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 2014

Direct and easy access to public transport information is an important factor for improving the satisfaction and experience of transport users. In the future, public transport information systems could be turned into personalized recommender systems which can help riders save time, make more effective decisions and avoid frustrating situations. In this paper, we present a predictive study of the mobility patterns of public transport users to lay the foundation for transport information systems with proactive capabilities. By making use of travel card data from a large population of bus riders, we describe algorithms that can anticipate bus stops accessed by individual riders to generate knowledge about future transport access patterns. To this end, we investigate and compare different prediction algorithms that can incorporate various influential factors on mobility in public transport networks, e.g., travel distance or travel hot spots. In our evaluation, we demonstrate that by combining personal and population-wide mobility patterns we can improve prediction accuracy, even with little knowledge of past behaviour of transport users.

2006

This paper proposes various measures regarding the variability of travel behaviours of transit users. The analyses are performed with smart card data collected over a ten months period. The variability in terms of boarding per day and new stops frequented with the days of travel on the transit network is examined. Data mining techniques are then used to classify days of travel according to the similarity of the boarding time periods. In this view, the use of two specific smart cards is examined in more details. These experiments first show that the behaviours of regular transit users evolve with time both in terms of transit stops frequented and time of boarding. Hence, variability of behaviours also changes for various user types.

2018

Variations in public transport system use can be observed from one user to another (interpersonal variations), but also within the behaviour of a same user over time (intrapersonal variations). This variability should be considered forecasting public transit demand. However, a better understanding of travel behaviours and tools to measure their variability (or regularity) are first required. This paper processes smart card data over a period of 51 weeks to investigate public transport passengers both interpersonal and intrapersonal variability using data mining methods. Several clustering algorithms based on different distances (Euclidian and non-Euclidian) are employed to build user typologies and thus highlight interpersonal variability. A typology of weeks is also created to reveal intrapersonal regularity, defined by the repetition of the same types of week in the user behaviour. Furthermore, indicators are provided to quantify the two types of variability. The results show that...

Journal of Transport Geography , 2021

The current work focuses on Quality of Service (QoS) of Public Transport (PT) attributes in urban tourist destinations. In particular, we aim to reveal which attributes are most significant for tourists prior to their arrival at their destination, as reflected in questions posted in TripAdvisor Question and Answer forums, a widely used social media platform. We used a data-mining method to classify questions into categories relevant to QoS, using a sample of 8,905 items posted between 2005 and 2018 in TripAdvisor forums for seven urban destinations in the United States and Western Europe. We found four PT-QoS attributes: Pricing and ticketing, Accessibility, Trip duration, and Service availability (hours of operation and frequency). These attributes have similar relative significance for all destinations, origins, seasons, and years we checked. Hence, they can help service operators and policymakers to understand tourists' preferences and to adjust PT services accordingly.