Donate to Science & Enterprise

S&E on Mastodon

S&E on LinkedIn

S&E on Flipboard

Please share Science & Enterprise

Math Model Identifies Network Source of Rumors, Epidemics

Pins on a map (CJ Sorg/Flickr)Computer scientists at the Swiss Federal Institute of Technology in Lausanne (EPFL) devised mathematical routines to dissect interactions in a network to uncover the source of epidemics and rumors, as well as criminal masterminds. Results of the research led by Pedro Pinto of EFPL’s Audiovisual Communications Laboratory appear today in the journal Physical Review Letters (paid subscription required).

Pinto and colleagues Patrick Thiran and Martin Vetterli wrote and tested an algorithm that makes it possible find the source of behavior in networks of all kinds based on the interactions of a small subset of those networks. As an example, Pinto uses a rumor about oneself circulating on the social network Facebook sent to some 500 people. “By looking at the messages received by just 15–20 of your friends, and taking into account the time factor,” says Pinto, “our algorithm can trace the path of that information back and find the source.”

The algorithm can also be applied to the behavior of non-technical interactions, as long as those interactions take the form of a network. Pinto applied the algorithm to find the source of a cholera epidemic in South Africa, using data collected by EPFL’s Ecohydrology Laboratory. “By modeling water networks, river networks, and human transport networks,” says Pinto, “we were able to find the spot where the first cases of infection appeared, by monitoring only a small fraction of the villages.”

The authors claim the techniques can be used as well to find the source of terror incidents or threats, using as an example, the telephone conversations believe to have occurred leading up to the terrorist attacks on 11 September 2001. The EPFL team reconstructed the message exchanges inside the 9/11 terrorist networks based on public news reports. Pinto says, “our system spit out the names of three potential suspects, one of whom was found to be the mastermind of the attacks, according to the official inquiry.”

While the case studies in the journal article use incidents in the past to validate the model, the authors say it can be applied prospectively as well, for example as a disease spreads in its early stages. “By carefully selecting points in the network to test,” says Pinto, “we could more rapidly detect the spread of an epidemic.” Another example of prospective use would be determining viral marketing strategies, where the algorithm would allow advertisers to identify influential blogs for the target audience and to understand how information in these blogs spreads through the online community.

Read more:

Image: C.J. Sorg/Flickr

*     *     *

1 comment to Math Model Identifies Network Source of Rumors, Epidemics