Abstract

Human mobility and social interactions have always been subjects of interest to many disciplines, including sociology, geography, and economics. These topics are key to address real-life problems like the design optimal transportation systems for goods and individuals, the control and prevention of epidemic spreading, and the forecast of social and economic consequences of migrations. Today, thanks to the availability of large and detailed datasets on people’s communications and whereabouts, we are in the position to test the validity of classical theories and models, improve them, and develop a deeper understanding of the underlying mechanisms governing our society. In this talk we introduce a decision-based stochastic model to estimate commuting flows, the radiation model, and discuss its connection with other spatial interaction models. We present some generalisations of the model, and analyse two scenarios in which we tested the model’s performance using GSM and GPS data. Finally, we discuss the contribution of spatial interaction models towards the understanding of the urbanisation process.