Certain programs exhibit revealing patterns in how they propagate
in the network.
We developed the SharkFin model [ASONAM 2013]
for the propagation rate of executable files on the
Internet.
This model incorporates three patterns observed in the field data:
(1) an exponential rise, leading to a spike;
(2) a power-law drop, after the peak propagation rate is achieved; and
(3) a daily periodicity pattern.
The parameters of these propagation patterns (e.g. the shape of the curve or the
velocity of propagation) likely correspond to the mechanism used to disseminate the file.
References
[ASONAM 2013] E. E. Papalexakis, T. Dumitraș, D. H. Chau, B. A. Prakash, and C. Faloutsos, “Spatio-temporal mining of software adoption and penetration,” in Advances in Social Networks Analysis and Mining (ASONAM), Niagara, Canada, 2013, pp. 878–885.
PDF