Geofast is developed by Dr Markus Esch and Connie Chan
under the supervision of Prof.
Vincent Blondel. Support is gratefuly acknowledged from Pierre
Deville, Dr Etienne Huens, Dr Zbigniew Smoreda, Prof. Jean-Paul
Donnay and Dr Thomas Aynaud.
Vincent D. Blondel, Gautier M. Krings, Isabelle Thomas, Regions and borders of mobile telephony in Belgium and around Brussels, Brussels Studies 42, ISSN 2031-0293, (13pp.), 2010.
Vincent D. Blondel, Jean-Loup Guillaume, Renaud Lambiotte, Etienne
Lefebvre, Fast
unfolding of communites in large networks, Journal of
Statistical Mechanics: Theory and Experiment, 1742-5468, P10008 (12
pp.), 2008.
To allow comparing arrondissements and communes with different numbers of customers Geofast normalizes the belgium data and presents per customer data.
France
The dataset covers the period between May 1, 2007 and October 31, 2007.
Only mobile communications
are used; text messages are removed. The location information used is
the position of the antennas. The maps therefore represent the amount
of traffic between antennas located in different "arrondissements" or
"départements" in France.
A description of the dataset can be obtained from:
Vincent Blondel, Pierre Deville, Frédéric Morlot, Zbigniew Smoreda,
Paul Van Dooren, Cezary Ziemlicki, Voice
on the border: do cellphones redraw the maps?, ParisTech Review,
2011.
See also the press coverage:
Le
mobile, reflet des frontières françaises, in "Le Monde", 17 décembre
2011.
In order to enable studying the temporal evolution of traffic, the french data has been scaled according to the minimum and maximum amount of traffic in the entire observation period per arrondissement (department) .
Ivory Coast
The dataset was collected for 150 days, from December 1, 2011 until
April
28, 2012. The original set contains 2.5
billion calls and text messages exchanged between around five million
users. The location information used is
the position of the antennas. This dataset is the one offered to the
selected participants to the "Data for Development" challenge organized
by Orange.
More information on the challenge and on the Ivory Coast dataset can be obtained
from http://www.d4d.orange.com/
Like the french data, the data from Ivory Cost has been scaled according to the minimum and maximum amount of traffic in the entire observation period. This allows studying the temporal evolution of traffic for a given department or region. Please note, that the Ivory Cost data is not complete. Around 25% of the calls are missing, which causes a sparse dataset on some days.