Database Reference
In-Depth Information
15.2.2 Interplay between Human Mobility and Social Networks
Recent advances on human mobility and social networks have turned the inter-
play between these two aspects into a crucial missing chapter in our understand-
ing of human behavior. To make progress in this direction requires large-scale
data that simultaneously capture dynamic information on individual movements
and social interactions. Thanks to the increasing availability of mobile phone
data sets and location-based online social networks (LBSN, see also Chapter
16
),
scientists have started to look into the questions of to what extent human mobility
patterns shape and impact our social ties, and how our social surroundings affect
where we go. The central hypothesis here is that social interactions increase with
physical proximity. Indeed, social links are often driven by spatial proximity,
from job- and family-imposed shared programs to joint involvement in various
social activities. These shared social foci and face-to-face interactions, repre-
sented as overlap in individuals' trajectories, are expected to have significant
impact on the structure of social networks. There are three lines of inquiry in
current literature: (1) geographic propinquity yields higher probability of form-
ing a tie; (2) overlap in trajectories predicts tie formation; (3) social environment
affects individual mobility.
Geographic Propinquity
The considerable influence of geographic distance on the formation, the evo-
lution, and the strength of friendships is probably rooted in the very nature
of our social brain. According to the anthropologist Robin Dunbar, there is a
physical cognitive limit in the number of strong ties the human brain is able
to manage, partly because it must be powered by a form of social grooming, a
time-consuming activity mainly based on geographical proximity and face-to-
face contact.
Recent analysis on Facebook and email data confirmed Dunbar's intuition,
showing that the volume of communications is inversely proportional to geo-
graphic distance and that the probability
P
(
d
) of having a friend at a certain
distance decreases following a sort of “gravitational law.” Although in the last
decades technology has contributed to reducing distances, proximity is still
important for the establishment of relevant relationships, breaking down the
illusion of living in “a global village”: a small world in which physical and
cultural distances vanish and where lifestyle become homogeneous.
In studying the social versus geography problem, data from LBSNs proved
to be very useful. Scellato et al. used information from both the social and loca-
tion components of several LBSNs to identify the relation between friendship
and geographic distance. They noticed a weak positive correlation between the
number of friends and their average distance, and observed that the socio-spatial
structure of the users cannot be explained by taking into account separately
