Network research published

New data indicator technique captures relevant information

Does ethnicity play a role in social network interactions? How can airlines develop a network of flights that balances a portfolio of short and long distance flights with economic viability?

A new technique developed by Ginestra Bianconi, a post-doc fellow in ICTP's Condensed Matter and Statistical Physics (CMSP) section, and colleagues Matteo Marsili of CMSP and Paolo Pin of the University of Siena, can help researchers assess whether certain network characteristics are essential or negligible for the structure of the network. It does this by helping to quantify the dependence of a network's structure on a given set of features.

"The technique defines a new indicator, different from any already available, that can capture relevant information present in the data where other methods have failed," said Dr. Bianconi. She described the method, which is based on network entropy measures, as a "statistical mechanics idea that can effectively be applied to complex networks and can be used for inferences."

An article on the technique has been published in the 15 June issue of the Proceedings of the National Academy of Sciences (PNAS).