Machine Learning for Smart Software Engineering Tools
Connections established by users of online social networks are influenced by mechanisms such as preferential attachment and triadic closure. Yet, recent research has found that geographic factors also constrain users: spatial proximity fosters the creation of online social ties. While the effect of space might need to be incorporated to these social mechanisms, it is not clear to which extent this is true and in which way this is best achieved.
To address these questions, we present a measurement study of the temporal evolution of an online location-based social network. We have collected longitudinal traces over 4 months, including information about when social links are created and which places are visited by users, as revealed by their mobile check-ins. Thanks to this fine-grained temporal information, we test and compare whether different probabilistic models can explain the observed data adopting an approach based on likelihood estimation, quantitatively comparing their statistical power to reproduce real events. We demonstrate that geographic distance plays an important role in the creation of new social connections: node degree and spatial distance can be combined in a gravitational attachment process that reproduces real traces. Instead, we find that links arising because of triadic closure, where users form new ties with friends of existing friends, and because of common focus, where connections arise among users visiting the same place, appear to be mainly driven by social factors.
We exploit our findings to describe a new model of network growth that combines spatial and social factors. We extensively evaluate our model and its variations, demonstrating that it is able to reproduce the social and spatial properties observed in our traces. Our results offer useful insights for systems that take advantage of the spatial properties of online social services.