WRM Temporal Network

Authors: Kemal Erdem, Dominika Kunc, Piotr Mazurek, Norbert Ropiak

Background

Model movement within the city
Bike usage patterns

Data

03.2019 - 12.2019
Wrocław Open Data

Preprocessing

ID uid bike_number start_time end_time rental_place return_place
20074 66240980 57809 2019-03-27 16:32:51 2019-03-27 16:33:35 Traugutta / Pułaskiego Traugutta / Pułaskiego
20075 66239215 57359 2019-03-27 16:15:15 2019-03-27 16:33:53 Wyszyńskiego / Szczytnicka Bardzka / Piękna
20076 66239036 57309 2019-03-27 16:13:14 2019-03-27 16:33:55 Śrubowa / Strzegomska Bardzka / Piękna
20077 66239414 57850 2019-03-27 16:17:12 2019-03-27 16:34:05 Śrubowa / Strzegomska Kościuszki / Pułaskiego
20078 66241086 57708 2019-03-27 16:34:09 2019-03-27 16:34:19 Traugutta / Pułaskiego Traugutta / Pułaskiego
20079 66240196 57782 2019-03-27 16:24:47 2019-03-27 16:34:21 Szewska / Kazimierza Wielkiego Komandorska / Sanocka

Data representation

  • Complete graph
  • Intervals - 15min
  • Nodes - stations
  • Edges - directed
  • Edge - weight

Metrics

  • Degree
  • In degree
  • Out degree
  • PageRank
  • Current Bikes Usage

Paths

$204^2$ total paths Dijkstra Algorithm

Display Graph

Optimization based on travel time between nodes

Interesting Findings

Teaching hours
Work travel

Interesting Findings - Psie Pole

How useful was
Network Science

  • Knowledge about Graph Theory
  • Deep understanding of measures
  • Paths finding
  • Temporal Networks

Further development

  • 2020 year data •
  • Expansion to other cities •
  • Realtime capacity data •
  • "Bike Tracing" •

Thanks

"There's no such thing as a stupid question!"

Authors: Kemal Erdem, Dominika Kunc, Piotr Mazurek, Norbert Ropiak
https://burnpiro.github.io/wod-bike-dataset-generator/
Acknowledgements: Icons designed by Creative_hat / Freepik