Mining Ticketing Logs for Usage Characterization with Nonnegative Matrix Factorization

Poussevin, Mickaël; Tonnelier, Emeric; Baskiotis, Nicolas; Guigue, Vincent; Gallinari, Patrick

doi:10.1007/978-3-319-29009-6_8

Cited by 14 publications

(9 citation statements)

References 21 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These works provide powerful solutions to predict traffic, but lack explanatory power. To capture spatial and temporal mobility patterns, existing efforts [10,23,27] use NMF to explain temporal patterns in daily life, such as commuting pattern that concentrates on mornings and afternoons, and explains the function of urban areas. In a recent work [30], a context-aware tensor decomposition is used to explain urban mobility over space and time using a tensor factorization approach.…”

Section: Related Workmentioning

confidence: 99%

Station-to-User Transfer Learning

Zhang

Pfoser

Züfle

2020

Proceedings of the 28th International Conference on Advances in Geographic Information Systems

View full text Add to dashboard Cite

Urban areas provide us with a treasure trove of available data capturing almost every aspect of a population's life. This work focuses on mobility data and how it will help improve our understanding of urban mobility patterns. Readily available and sizable farecard data captures trips in a public transportation network. However, such data typically lacks temporal signatures and as such the task of inferring trip semantics, station function, and user clustering is quite challenging. While existing approaches either focus on stationlevel or user-level signals only, we propose a Station-to-User (S2U) transfer learning framework, which augments user-level learning with shared temporal patterns learned from station-level signals. Our framework is based on a novel, so-called "Tidal-Regularized Non-negative Matrix Factorization" method, which incorporates a-priori tidal traffic patterns in generic Non-negative Matrix Factorization. To evaluate our model performance, a user clustering stability test based on the classical Rand Index is introduced as a metric to benchmark different unsupervised learning models. Using this metric, quantitative evaluations on three real-world datasets show that S2U outperforms two baselines methods by 7 − 21%. We also provide a qualitative analysis of the user clustering and station functions for the Washington D.C. metro and show how S2U can support spatiotemporal urban analytics.

show abstract

Section: Related Workmentioning

confidence: 99%

Station-to-User Transfer Learning

Zhang

Pfoser

Züfle

2020

Proceedings of the 28th International Conference on Advances in Geographic Information Systems

View full text Add to dashboard Cite

show abstract

“…However, some more advanced machine learning methods have also been developed in recent works. For instance, nonnegative matrix factorization (NMF) is used in [12] to discover a dictionary of behavioral atoms to describe passengers based on their subway journey transactions. The distribution of these atoms over the stations is then used to conduct multi-scale clustering and retrieve groups of stations with similar behavior.…”

Section: Related Workmentioning

confidence: 99%

A mixture model clustering approach for temporal passenger pattern characterization in public transport

Briand¹,

Côme²,

Mahrsi³

et al. 2016

Int J Data Sci Anal

View full text Add to dashboard Cite

Smartcard data provide a great number of information that are increasingly used nowadays. In the field of transport, they offer the opportunity to study passenger behavior, leading to a better knowledge of public transit demand and thereby granting the transport operators the ability to adapt their transport offer and services accordingly, both in space and in time. In particular, an accurate characterization of mobility patterns using data mining approaches has a very strong interest for transport planning purposes. This paper aims to propose a two-level generative mixture model that partitions passengers according to their temporal profiles. Using the timestamps of the passengers' transactions in the public transport network, the first level models the passengers partitioning into a reduced set of clusters, whereas the second level captures how the trips made by each cluster of passengers are distributed over time. The proposed approach is applied on real ticketing data collected from the urban transport network of Rennes Métropole (France). The obtained results show that different passenger profiles can be discovered, thus highlighting several patterns of transport demand. The crossing of the clustering results with smartcard fare types as well as city characteristics such as academic centralities is also conducted in order to identify the close link B Anne-Sarah Briand

show abstract

“…Depending on the targeted objective, a distinction is usually made between unsupervised methods with an exploratory purpose and supervised methods with a prediction or classification objective. Various clustering approaches have been developed to highlight group structures in user routines (Lathia et al 2013;Briand et al 2016;He, Agard, and Trépanier 2020) or in the use of transport systems (Poussevin et al 2015;El Mahrsi et al 2016). Principal Component Analysis (PCA) is another unsupervised method used by Luo, Cats, and van Lint (2017) to provide insight into the underlying structure of flow dynamics within a metro network.…”

Section: Introductionmentioning

confidence: 99%

What can we learn from 9 years of ticketing data at a major transport hub? A structural time series decomposition

Nailly

Côme

Samé

et al. 2021

Transportmetrica A: Transport Science

View full text Add to dashboard Cite

Mobility demand analysis is increasingly based on smart card data, that are generally aggregated into time series describing the volume of riders along time. These series present patterns resulting from multiple external factors. This paper investigates the problem of decomposing daily ridership data collected at a multimodal transportation hub. The analysis is based on structural time series models that decompose the series into unobserved components. The aim of the decomposition is to highlight the impact of long-term factors, such as trend or seasonality, and exogenous factors such as maintenance work or unanticipated events such as strikes or the COVID-19 health crisis. We focus our analysis on incoming flows of passengers to two transport lines known to be complementary in the Parisian public transport network. The available ridership data allows analysis over both long-term and short-term time horizons including significant events that have impacted people's mobility in the Paris region.

show abstract

Mining Ticketing Logs for Usage Characterization with Nonnegative Matrix Factorization

Cited by 14 publications

References 21 publications

Station-to-User Transfer Learning

Station-to-User Transfer Learning

A mixture model clustering approach for temporal passenger pattern characterization in public transport

What can we learn from 9 years of ticketing data at a major transport hub? A structural time series decomposition

Contact Info

Product

Resources

About