New research tracks Twitter social media data to predict large social events

Google+ Pinterest LinkedIn Tumblr +

Northeastern’s Alessandro Vespig­nani, Stern­berg Family Dis­tin­guished Uni­ver­sity Pro­fessor of physics, com­puter sci­ence, and health sci­ences, has teamed up with an inter­dis­ci­pli­nary group of sci­en­tists to develop an inno­v­a­tive method to map how tweets about large- scale social events spread.

Using mas­sive twitter datasets and sophis­ti­cated quan­ti­tative mea­sures, it tracks how infor­ma­tion about polit­ical protests, large busi­ness acqui­si­tions, and other “col­lec­tive phe­nomena” gather momentum, peak, and fall over time, from city to city, and where the impetus comes from for that trajectory.

The find­ings, pub­lished Friday in the journal Science Advances, is only a first step, notes coau­thor Nicola Perra, a former research asso­ciate at Northeastern’s Net­work Sci­ence Insti­tute. But knowing the char­ac­ter­is­tics of that buildup could, in the future, enable us to pre­pare ahead of time for unde­sir­able reper­cus­sions from such events, with impli­ca­tions for crises from earth­quakes to power- grid failures.

“A lot of people have ana­lyzed social media in terms of the volume of tweets regarding par­tic­ular phe­nomena such as the Arab Spring,” says Vespig­nani, who is also the director of the Net­work Sci­ence Insti­tute. “What we are trying to under­stand is the pres­ence of pre­cur­sors: Can we find a signal in the flow of infor­ma­tion that will tell us some­thing big is about to happen? That’s the multimillion- dollar question.”

In an inter­dis­ci­pli­nary leap, the researchers turned to net­work mod­eling in neu­ro­science to con­duct the study. “For the brain we map based on phys­i­ology, and for social aggre­gates, like those in this paper, we map on geog­raphy,” says Vespignani.

In neu­ro­science net­work mod­eling, the “nodes,” or cen­ters of activity, are func­tional brain areas–say, the motor cortex, which is respon­sible for movement–and the “links” con­necting the nodes are neural cir­cuits. For example, the cir­cuits con­necting the motor cortex to the audi­tory cortex, which is respon­sible for hearing, trace a neural pathway, or “link,” that enables us to tap our foot to a beat and even dance.

In this new, social- events study, the nodes are cities–for example, Madrid and Barcelona in the researchers’ analysis of twitter trans­mis­sion during the 2011 Spanish anti- austerity movement–and the links are the path­ways the tweets take over time.

Con­sider the Spanish protest, which later sparked the Occupy Wall St. move­ment in the U.S. The tweets gained in volume and inten­sity until, says Vespig­nani, they reached a “social tip­ping point of col­lec­tive phe­nom­enon” on May 20, 2011. “You create a system that starts from a few nodes that then drive others, and so on, until every­body is talking to every­body else in a full coor­di­na­tion of the infor­ma­tion,” he explains.

The quan­ti­ta­tive iden­ti­fi­ca­tion of those dri­vers sets this new method apart from other approaches to tracking social media, says Perra, who is now a senior lec­turer at London’s Uni­ver­sity of Green­wich. “It enables us to under­stand which city is dri­ving the con­ver­sa­tion when and to char­ac­terize the dynamics of the spread.”

“Before you can develop a method to pre­dict future events,” he adds, “you need a quan­ti­ta­tive under­standing of the com­mu­nica­tion pat­terns that shaped past events.”

Laying the ground­work for pre­dic­tive studies is what Vespig­nani and his col­leagues are attempting to do with this analysis of five major social events: the 2011 protest in Spain; the 2013 protest in Brazil, known as the “Brazilian Autumn”; the release of a Hol­ly­wood block­buster in 2012; and Google’s acqui­si­tion of Motorola in 2014.

“Everyone wants to pre­dict when the next big event is going to be, what will trend in the future,” says Perra. “We are, as a research com­mu­nity, in the early stages of under­standing this type of phe­nomena. There is very little under­standing of even past events, so we are very far from pre­dic­tion. But in the future our find­ings may lead us to that.”

Share.