Suchergebnisse

Nowadays, the striking proliferation of Big Data and the new scientific tools provided by the emerging field of Data Science finally offers the opportunity to realize a longstanding dream of scientists and policy makers: drawing a comprehensive picture of human social behavior. Big Data, indeed, hide a huge amount of predictive power, which can be exploited by governments and companies to unveiling and understanding the complexity underlying our society. In the present thesis we propose a multidimensional study of human social behavior, aimed to understand how the social network, the mobility behavior and the economic wellness of people in a big European country are connected to each other. To this end we exploit the access to a big mobile phone dataset provided by the French company Orange. Thanks to Big Data management tools like Hadoop, we computed several individual measures, each describing different aspects of the social or mobility behavior of individuals. Our analysis revealed very interesting and striking findings. Firstly, the observation at neighborhood level of the three biggest cities in France (Paris, Marseille, Lyon) uncovers a very strong correlation between the social diversification and the mobile predictability. People who equally diversify their calls over the social contacts tend to have a more erratic mobile behavior. Moreover, an even more interesting relationships emerged between the mobile behavior and the economic status: mobility diversity is strongly correlated with some indexes of economic wellness. These striking results leaves no room for doubt: the greater the diversification of the mobile behavior of people within a territory, the higher their economic prosperity. Such findings open interesting future perspective about the study of human social behavior. New statistical indexes can be defined which rely on mobile phone data to describe and predict the actual and future economic health of a territory. Tell me how people interact and move, and I will tell you how they live.