All posts by Ying Wang

CNetS research associates present epidemic visualization and modeling tools at New Orleans supercomputing conference

Bruno Gonçalves
Bruno Gonçalves

Duygu Balcan and Bruno Gonçalves represented the Pervasive Technology Institute and the Center for Complex Networks and Systems at the 23rd Supercomputing Conference in New Orleans. At the conference they presented the Global Epidemic and Mobility (GLEaM) computational tool and the novel visualization component “Epidemic Planet” to researchers from the public and private sectors around the world.

The novel Epidemic Planet visualization, developed in collaboration with the ISI foundation, includes a touch-screen interface that allows users to set parameters for two separate pandemic outbreaks, and another screen that visualizes the pandemic spread in each case as an overlay on one of two global maps. Users can parametrize pandemic scenarios by country of origin, level of infectiousness, time of year, and other factors that define different disease scenarios.

Duygu Balcan
Duygu Balcan

The application is in continuous development, and has been shown at the Science Gallery at the Trinity College inDublin, the Edinburgh International Science Festival, the Science Beyond Fiction event at the EU Parliament in Strasbourg, the CosmoCaixa Science Museum in Barcelona, and more.


LaNeT-vi visualization featured on the cover of Nature Physics

Cover of Nature Physics Nov. Edition

LaNeT-vi, a program that represents large-scale networks in two-dimensions, was used to create the cover image for the November edition of Nature Physics. The featured image highlights several nodes in the foreground that stand out from the rest of the network, highlighting the main thrust of the journal’s cover story; the most efficient spreaders in a network are not necessarily the most connected or central. Instead, efficient spreading correlates with spreader location within the core of the network as determined by the k-shell decomposition analysis.

Determining what makes an efficient spreader in a network is essential for optimizing network efficiency and the deployment of resources. As a publicly-available online tool, LaNeT-vi assists in this effort by allowing users to upload their own networks and receive original renderings of their information based on the k-core decomposition.

LaNeT-vi was developed in-house at the School of Informatics and Computing at Indiana University in collaboration with the CNRS, France and the CONICET in Argentina.

“Clouds are not spheres, mountains are not cones.”

Initial image of a zoom sequence: Mandelbrot set with continuously colored environment (Created by Wolfgang Beyer with the program Ultra Fractal 3)

Benoit B. Mandelbrot, the ‘Father of Fractals’, died Thursday, October 14 in Cambridge, Mass. He was 85. His passing marks a great loss for science.

Fractal mathematics has provided us with the key to accessing the geometric complexity of nature by assisting us in the decoding and ordering of chaotic and disordered systems. The conceptual power of fractal mathematics is surpassed only by its stunning visual beauty. Benoit was a maverick mathematician who constantly made forays into both physics and economics where his impact has been profound and is difficult to quantify even today.

He was never afraid to go against mainstream conventions and never settled on performing incremental scholarly work. Instead, Benoit aimed at changing paradigms, and indeed he was one of few who ever managed to do so.

In his famous 1982 book, he wrote, “Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.” Beyond our course-grained and efficient conceptions of nature’s geometry lies a reality that is even more fascinating, complex, and sublime than we had imagined.

Thank you Benoit; you have changed the way we look at the world around us.

CNetS Associate Professor Johan Bollen interviewed on CNBC and CNN for his work on sentiment tracking via Twitter

Johan Bollen discussed his findings that the public’s collective mood as expressed on Twitter correlates in a surprising way to the performance of the Dow Jones Industrial Average today on CNBC’s Squawk on the Street and CNN International’s Quest Means Business. Bollen explained that contrary to expectations, the public’s overall anxiety appears to predict DJIA closing values three to four days in advance, rather than following from DJIA performance.

The interview draws from a paper authored by Bollen and colleagues called “Twitter Mood Predicts the Stock Market” wherein the researchers presented their findings after applying the Google-Profile of Mood States (GPOMS) algorithm to 9.7 million tweets between March and December 2008. The algorithm measures public mood by analyzing the textual content of Tweets before indexing that content into one of six mood states. One of the mood states “calmness” was demonstrated to be significantly predictive of market performance.

 

New article published in Physical Review Letters puts forth new model of online popularity dynamics

Online popularity can be thought of as analogous to an earthquake; it is sudden, unpredictable, and the effects are severe. While shifts in online popularity are not inherently destructive – consider the unprecedented magnitude of online giving via Twitter following the disaster in Haiti – they indicate radical swings in society’s collective attention. Given the increasingly profound effect that large-scale opinion formation has on important phenomena like public policy, culture, and advertising profits, understanding this behavior is essential to understanding how the world operates.

In this paper by Ratkiewicz and colleagues, the authors put forth a web-wide analysis that includes large-scale data sets of the online behaviors of millions of people. The paper offers a novel model that is is capable of reproducing all of the observed dynamics of online popularity through a mechanism that causes sudden, nonlinear bursts of collective attention. These results have been mentioned in the APS and PhysOrg websites.

Article in Nature on the plenitude of scientific performance metrics features the research of Bollen and Vespignani

The relatively new field of bibliometrics has experienced an explosion of research as scientists become more interested in developing metrics that can accurately measure scientists’ performance. The common but naive practice of tallying the number of journal citations accumulated by researchers has serious limitations insofar as many salient factors like the weight of a citation as a function of a journal or database’s popularity, how well an article integrates with contemporaneous research, and individual productivity are not taken into account.

The article discusses Bollen’s concern that the scramble to uncover new metrics and combinations of them has obscured an equal need to define the concepts under measurement more rigidly. It also addresses an approach taken by Vespignani and colleagues to apply the concept of weighted citations to develop a network of over 400,000 papers published over 100 years in order to demonstrate the variable influence scientists have over the scientific community. Read more…