The social network's structure optimizes the sharing of information in a way that Twitter's hierarchical setup does not.
PROBLEM: The goal of a social network, we constantly try to convince our sharing-everything-on-the-Internet-averse friends, is to connect to others in a way that enhances -- not replaces -- the real-world communities in which we exist. But if we were to zoom out and look at everything happening on Facebook beyond the limited vantage point of our newsfeeds, would we see that this is actually the case?
- Domestic Violence Increases After Major Sporting Events
- People Who Exercise Have Larger Brains Later in Life
- People Living in Poverty Are Twice as Likely to Be Depressed
METHODOLOGY: Sampling something as large and amorphous as Facebook is no simple task. Emilio Ferrara, of both Indiana University and Italy's University of Messina, used sophisticated data-mining to identify public profiles, developed complex algorithms to track users' interactions with other users, and mapping the networks that emerged. Users remained anonymous -- Ferrara reduced us all to numbers in order to understand our interactions from a statistical viewpoint.
RESULTS: Ferrara discovered that, without intending to do so, Facebook users are driven by the network's very structure to group themselves into a large number of small communities, all joined together into the monolith we call Facebook by their social connections. Large groups containing a majority of users are a rarity, but that doesn't mean that each group of friends exists in its own bubble -- instead, the smaller communities are highly interconnected in a way that two people with nothing in common are often connected by short, easily traversable paths of communication.
IMPLICATIONS: The structure discovered to exist on Facebook fits nicely into a social theory called the "strength of weak ties." Basically, it works by saying that if Sigmund Freud is connected to both Hippocrates and Dr. Oz -- they've both accepted his friend requests -- then Hippocrates and Dr. Oz almost certainly share a connection -- apart from their connection to Freud -- as well. They aren't Facebook friends, and they probably don't have much in common, but their personal social networks slightly overlap.
These loose connections, goes the theory, end up making communication more efficient. Because when something happens that passes by Dr. Oz's radar, his good friend Freud isn't likely to have heard about it either. He's much more likely to find out what's going on from Hippocrates, because Hippocrates knows all sorts of people who Dr. Oz and Freud do not. Weak ties like these, which cropped up everywhere in Ferrara's model, are what connect small, disparate networks to one another.
On Twitter, one person, the Followed, spreads information to his or her followers. Those followers can in turn share with their followers, but the flow of information always follows a top-down hierarchy. Weak ties abound here, too, but not, argues Ferrara, to the same extent as they do on Facebook. The communities built through Facebook, he says, are more true-to-life.
The full study, "A large-scale community structure analysis in Facebook," is published in the journal EJP Data Science .