It’s a rather unsur­prising idea: Humans do things in bursts of activity. “We do not do things uni­formly,” said Albert-​​László Barabási, a Dis­tin­guished Pro­fessor of Physics with joint appoint­ments in the Col­lege of Sci­ence and the Col­lege of Com­puter and Infor­ma­tion Sci­ence and founding director of Northeastern’s world-​​leading Center for Com­plex Net­work Research.

Instead, he said, each one of our actions is cor­re­lated to those that came before. For instance, per­haps we send 16 emails in a half-​​hour period and then surf the Web for an hour. Or maybe we make a few calls in the after­noon and then set the phone down for the rest of the evening.

Though it may be a sub­con­scious side-​​effect of evo­lu­tion or simply con­ve­nience, our bursty behavior allows net­work sci­en­tists like Barabási to pre­dict how we will act in the future and, by exten­sion, how the prod­ucts of our bursty work — our money, our tweets or even our ideas — will travel through society.

But these pre­dic­tions depend on robust math­e­mat­ical algo­rithms. Until recently, the net­work sci­en­tists’ stan­dard tools tended to “give rise to pos­i­tive cor­re­la­tions, even though the events were totally inde­pen­dent,“ said Barabási. For pre­dom­i­nantly random sys­tems, these models work fine. Not so with bursty systems.

In a recent paper pub­lished in the journal Nature Sci­en­tific Reports, Barabási and his col­leagues report a new model capable of detecting bursti­ness more accurately.

Márton Karsai, the paper’s first author — and a researcher who will join the lab of physics pro­fessor Alessandro Vespig­nani this fall — said that up until now, “we have pre­sumed that these sys­tems can be inter­preted as mech­a­nisms where the con­sec­u­tive events are not inde­pen­dent from each other and are cor­re­lated.” Now they can con­firm that pre­sump­tion mathematically.

Bursti­ness is not exclu­sive to the human social net­work. Other ani­mals also behave this way. Our neu­rons fire in bursts. Even earth­quakes follow the pat­tern, with long periods of silence fol­lowed by a col­lec­tion of tremors. The new model reveals a set of prop­er­ties uni­versal to all of these sys­tems, despite the fact that they are driven by extremely dif­ferent mechanisms.

The common fea­ture shared by these sys­tems is that beyond being bursty, the bursty events are not evolving in pairs but more likely in longer trains,” Karsai said. That is, the var­ious indi­vidual events in a burst are all cor­re­lated; the rela­tion­ship does not exist only between those events that follow one another in a sequence.

The new algo­rithm, Barabási said, “offers fur­ther mod­eling and con­cep­tual chal­lenges for the field.” Karsai added that they have already begun to move in this direc­tion, having pub­lished another paper using this method to answer more spe­cific ques­tions about human com­mu­ni­ca­tion behavior.

View selected pub­li­ca­tions of Albert-​​László Barabási in IRis, Northeastern’s digital archive.