Understanding these networks can aid researchers in devising plan for devastating natural disasters, such
as the eruption of the Eyjafjallaj¨okull volcano in 2010 [10] or Ebola outbreak [4]. While real networks
are insightful, they are usually hard to obtain (such as obtaining repetitive samples of the same type of
network to create temporal networks; data collected to create networks may contain personable identifiable
information; or the sampled data may be at the wrong scale many times needing to size it up).

Researchers desire methodologies to create tunable synthetic networks that mimic the real ones.
These will then allow the researchers to change the tuning parameters to create different scales of networks
that have similar properties to the observed in the real networks. As the networks change with time, so
too do the graphs that represent them, and therefore their properties, such as communities and network
vulnerabilities.

Your goal is to develop accurate, robust, and tunable temporal network models with communities, that
can be tailored primarily for cyber applications. That is create synthetic networks that:
1. are temporal,
2. have communities (GUI for visualization: Gephi or http://www.cfinder.org/),
3. vary parameters (to get different scales, community structure, or network size for example),
4. have similar properties to real ones (matching the properties of the real one when the synthetic network
is at the same scale as the real network).
Most existing synthetic models describing the evolution of communities in dynamic networks use spectral
graph theory, Dirichlet process mixture model, stochastic block model, or component structure
model.

Exploit existing methodologies and introduce innovative approaches to create synthetic networks model-
ing time-evolving cyber graphs that change the community structure and adapt as changes in the network
occur. In creating these networks, consider probabilistic communities that are born, grow, decay, split,
merge, or die within specified ranges of speeds, sizes, and noise levels; an example of which can be seen
in the Figure below.  Ultimately, introduce a model that represents real cyber networks of interest to DoD stakeholders.