Authors: Isabelle Thomas*, National Fund for Scientific Research, Arnaud Adam, Université catholique de Louvain, Olivier Finance, Université catholique de Louvain
Topics: Transportation Geography, Quantitative Methods, Economic Geography
Keywords: Transport geography, ICT data, Belgium
Session Type: Paper
Start / End Time: 3:20 PM / 5:00 PM
Room: Napoleon C2, Sheraton 3rd Floor
Presentation File: No File Uploaded
By means of several datasets related to movements of people and goods (freight) in Belgium, we discuss the uses and misuses of ICT data in (transport) geography. More particularly we point out the dilemma between (1) the growing availability of large data sets for revealing the pulses of “smart” urban, periurban and rural environments (high versus low frequency geographies) and (2) the associated (in-)capacity of increasing geographical theoretical/long term knowledge. More particularly, we here use two train schedules applications on smartphones, fluxes of taxes between logistic companies and GPS tax trackers on trucks and discuss them critically. Places/people that are tightly related are optimally clustered by means of a community detection algorithm (the Louvain Method) leading to new partitions of Belgium. We obtain a mosaic of “interaction fields” that are interpreted in terms of methodological choices, human and urban geography/economics as well as Belgian political dilemmas. They give the opportunity to remind that researchers should always control the impact of their methodological choices on their final results and that each type of data leads to a different geographical partitioning and hence to different theoretical and political interpretation/decisions. Particular attention is put on the importance of the regional/linguistic borders.