In this paper, we analyze public transport passenger movement data to detect typical patterns. The initial data consists of smart card transactions made upon entering public transport, collected over the course of two weeks in Saint Petersburg, a city with a population of 5 million. As a result of the study, we detected 5 classes of typical passenger movement between home and work, with the scale of one day. Each class, in turn, was clusterized in accordance with the temporal habits of passengers. Heat maps were used to demonstrate clusterization results. The results obtained in the paper can be used to optimize the transport network of the city being studied, and the approach itself, based on clusterization algorithms and using heat maps to visualize the results, can be applied to analyze public transport passenger movement in other cities.
Original languageEnglish
Title of host publicationTrends and Innovations in Information Systems and Technologies
PublisherSpringer Nature
Pages555-563
Volume2
ISBN (Electronic)978-3-030-45691-7
ISBN (Print)978-3-030-45690-0
DOIs
StatePublished - 1 Jan 2020
Externally publishedYes
EventTrends and Innovations in Information Systems and Technologies - Budva, Montenegro
Duration: 7 Apr 202010 Apr 2020

Publication series

NameAdvances in Intelligent Systems and Computing
PublisherSpringer Nature
Volume1160
ISSN (Print)2194-5357

Conference

ConferenceTrends and Innovations in Information Systems and Technologies
Country/TerritoryMontenegro
CityBudva
Period7/04/2010/04/20

    Research areas

  • Multimodal trips, Pattern mining, Public transport, Urban transit system

ID: 103098236