Boullier, Dominique (2023) « New era for commensurable comparative urban research ? Machine Learning and/or propagations », in Le Galès P. & Robinson J., Routledge Handbook of Comparative Global Urban Studies, 246-259.
This paper discusses the methodological choices made available by high level digital methods (Machine Learning) in comparative urban studies. A review of papers from the journal Urban Studies is taken as a sample to demonstrate the potential and the risks of a non controlled use of Big Data. The amplification of usual standpoints should be challenged by the emergence of traceability as a new statistical quality that makes possible the follow-up of propagations in urban experiences and to compare them systematically.