Database Reference
In-Depth Information
The spreading of biological and mobile phone viruses is another context in
which geo-social data could be useful, because epidemics are also determined
by the structure of social and contact networks within the population, and human
mobility patterns. The mathematical modeling of infectious diseases must take
into account travel patterns within a city or the entire world, and accurately shape
the underlying contact network depending on the nature and the infectiousness of
the pathogen. For example, with highly contagious diseases (e.g., transmission
based on coughs and sneezes) the contact network will include any pair of
people who sat together in the same place. For a disease requiring close contact
(e.g., sexually transmitted disease), the contact network will be much sparser.
Similar distinctions arise in the computer virus context, where malware infecting
computers across the Internet will have a much broader contact network than
one that spreads by short-range wireless communication between nearby mobile
devices. Depending on the case, a contact network based on co-location in
a place or the explicit social network could be inferred by using geo-social
data from geo-social networks, geo-tagged photo websites, or geo-referenced
microblogging websites. Some research in this direction has been conducted,
but it is far away from being exhaustive.
Mobility patterns of a population can be extracted by using check-in trajec-
tories of users, in order to define the epidemic model or to perform a simulation
scenario. A very fascinating application is the development of mobility models
and routing algorithms for the so-called opportunistic networks. They are a new
paradigm of computation in which there is no fixed infrastructure, and mobility
is exploited as an opportunity to deliver data among disconnected parts of a
network. When a node has data to transfer to another node, and no network path
exists between the sender and the receiver, any possible encountered mobile
device represents an opportunity to forward and carry them until encountering
another node deemed more suitable to bring the message to the final destination.
Both in the design of routing algorithms and in the evaluation of them, a promis-
ing approach is that of incorporating the spatial dimension into a model based
on time-varying social graphs. Geo-social data are clearly the most appropriate
and useful tool in this context because thus provide at the same time all three
dimensions of human movements: spatial, temporal, and social dimensions. In
addition to this, explicit social relationships from online social networks can be
incorporated to better design protocols that are able to learn the social network
of users, for example in order to exploit the role of hubs (users with the highest
number of contacts) in the dissemination process, or to predict new friendships
and contact opportunities.
These examples are, of course, not exhaustive. They just give a hint of
the possible uses of CGI data in several different context scenarios. The tech-
nology and the application are moving and changing so fast that some very
unexpected and innovative applications can be developed even in the next few
months.
