Skip to Content

Researchers Peek at the Structure of the Viral Internet

Microsoft shows how it tracked the spread of more than 1.4 billion tweets.

A video or article or meme “going viral” is one of the most overused terms on the Internet. The term “virality” is taken from the field of epidemiology, where scientists using computer models attempt to map the spread of a disease. On the Internet, however, we have viral marketers. These are folks who can talk a good game— they even have mathematical equations with borrowed ideas like the virality coefficient—but very little in the way to measure these in a rigorous way. Mostly, we resort to saying something has gone viral simply if it gets a lot of views on YouTube, is trending on Twitter, or spawns a mutated meme. And, of course, if your friends already know about it when you bring it up. Oh, THAT, video. 

An example analysis from Microsoft’s Viral Search.

At Microsoft Research’s annual technology demo day this week, researchers showed off a tool called Viral Search that attempts to measure virality in its more literal sense. That means not overall traffic over time, but the mechanics by which it passes from person to person over many generations. 

The software looked at 1.4 billion Tweets over the course of a year and produced a branching tree that shows how links spread, based on who follows whom and the sequence of retweets. From this, researcher Jake Hofman and a team created a new kind of Web metric that measures the average distance on the tree between any two people who tweeted the story. More distance means more virality, since there were more paths by which people passed on a link. Something can be relatively “viral,” like an online petition that people sign and often pass on, without getting near Gangham Style levels of hits. The opposite of virality might be a random tweet from Justin Bieber that got a lot of initial hits because he’s Justin Bieber, who has 35 million followers, but died off pretty quickly, because how many other people really care what he’s doing on his family vacation. Most story trees on Twitter were pretty small, with fewer than two “generations” before they died out, the analysis found. 

The researchers have packaged these large computations into a neat search and visualization tool, and the conclusion that true virality is rare isn’t astonishing or new.  But what’s most interesting were these ideas under the hood. In epidemiology, tracing a germ from person to person so finely is usually not possible. On Twitter it is. That’s the biggest weakness, however. Since Twitter is a social network that is meant to be public, there aren’t actually many other parts of even the Internet world where doing something like this would be possible at any meaningful scale, Hofman says. You can watch a video of the demo here

Keep Reading

Most Popular

The miracle molecule that could treat brain injuries and boost your fading memory

Discovered more than a decade ago, a remarkable compound shows promise in treating everything from Alzheimer’s to brain injuries—and it just might improve your cognitive abilities.

wet market selling fish
wet market selling fish

This scientist now believes covid started in Wuhan’s wet market. Here’s why.

How a veteran virologist found fresh evidence to back up the theory that covid jumped from animals to humans in a notorious Chinese market—rather than emerged from a lab leak.

Conceptual illustration showing a file folder with the China flag and various papers flying out of it
Conceptual illustration showing a file folder with the China flag and various papers flying out of it

The US crackdown on Chinese economic espionage is a mess. We have the data to show it.

The US government’s China Initiative sought to protect national security. In the most comprehensive analysis of cases to date, MIT Technology Review reveals how far it has strayed from its goals.

conceptual illustration showing various women's faces being scanned
conceptual illustration showing various women's faces being scanned

A horrifying new AI app swaps women into porn videos with a click

Deepfake researchers have long feared the day this would arrive.

Stay connected

Illustration by Rose WongIllustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.