With the advent of pervasive location services researchers have gained an unprecedented access to records of human activity in space and time. We analyzed geo-located Twitter messages in order to uncover global patterns of human mobility. Based on a dataset of about one billion tweets recorded in 2012, we analyzed the spatial behavior of Twitter users globally and their international travels.
We looked at various types of human mobility, but in particular at the following:
Our analysis of the community structure of the Twitter mobility network reveals spatially cohesive regions that follow the regional division of the world. The results were validated with global tourism statistics and based on the comparison to common mobility models such e.g. the gravity law. The project confirmed that Twitter is exceptionally useful for understanding and quantifying global mobility patterns.
Further details can be found in the publication 'Geo-located Twitter as proxy for global mobility patterns'.