Clive Thompson on Real-World Social Networks vs. Facebook

Moderators: Elvis, DrVolin, Jeff

Clive Thompson on Real-World Social Networks vs. Facebook

Postby §ê¢rꆧ » Wed Aug 06, 2008 12:37 pm

This whole idea of "reality mining" seemed sufficiently creepy to post about here...

Clive Thompson on Real-World Social Networks vs. Facebook 'Friends'

Benjamin Waber has a grim piece of news for managers and CEOs: You're out of the loop.

Waber, a PhD student in MIT's Human Dynamics Group, studies the way groups interact socially — based on who's talking to whom. But unlike most social scientists, who simply ask people about their behavior, Waber and his colleagues measure it. They outfit employees with special badges that work with base stations to log all conversations between employees, including location and duration. With this data, Waber's team can plot exactly how information flows inside a firm.

Almost every time he analyzes a group, Waber discovers that the super-connector — the crucial person who routes news among team members — isn't the manager. "The manager is almost always peripheral," Waber says. "It's some random guy." And that person is usually overworked and overstressed. He isn't given enough support to fulfill his role, because nobody in the firm knows he's doing it in the first place. If you study the org chart, the higher-ups are in control. But if you study reality, those same managers barely know what's going on.

This type of research has evolved into a new field called reality mining. By tracking people using location-aware devices like mobile phones or electronic badges, scientists are revolutionizing our understanding of how social networks function.

Of course, we think we know how they work. We've all become addicted to some combination of email and LinkedIn and Facebook and blogs, and at the click of a button we can pinpoint our online friends, right? But once you step away from your computer, Waber and other reality miners have found, the real world often works in ways that are quite different from the virtual one.

On the Web, the best way to solve a problem is to engage an extensive network; the person who provides information, advice, or answers is often someone you know only vaguely — a weak link.

In the face-to-face world, though, Waber says, groups are more productive when the team members know each other well, sharing extremely strong links. That's because face-to-face teamwork requires intimacy, he says, and "when you're among friends you can really capitalize on preexisting protocols" — nods, grunts, in-jokes — for talking and listening.

Reality mining can also spot when a group is in a groove. Sandy Pentland, the MIT professor who heads up the lab where Waber works, has discovered that highly creative teams socialize in a "pulsing star" pattern: They fan out to gather information, then regroup. "People explore during the day," Pentland says, "and then later get very tight and inbred, with everybody talking to everybody."

If you have enough data about commonplace conversations, you can even predict when those conversations are going to take place. Working with Pentland, Nathan Eagle tracked the physical interactions of 100 MIT students over an academic year, using their cell phones. After a few months, Eagle could deduce likely future meetings with impressive accuracy. "So if we know that," he says, "why not design our calendars to sync up?"

But this isn't just about understanding reality. It's about tweaking it. When Waber examines company-wide communications, he can spot inefficiencies — two employees who don't know each other. Introduce them over coffee and — presto! — the office metabolism accelerates. The technical term for this is tightening a network, and Waber is trying it out at several firms around the world.

The scary part of reality mining, as everyone involved readily admits, is that it's a potential privacy nightmare. Do you really want your boss gathering this much data about your daily activities? Or imagine this stuff in the hands of direct-marketing types. Or law enforcement.

Still, the benefits might outweigh the risks, particularly at the public-policy level. Mining companies is cool; mining countries could yield lifesaving info. Eagle is currently analyzing 12 billion anonymized calls placed during one month in the UK. With that much data, he hopes to better understand human mixing patterns, which could help predict the spread of disease outbreaks, social trends, and other hidden phenomena.

We've learned to map our virtual lives. Now it's time to map the real ones, too.

Code: Select all
User avatar
Posts: 1197
Joined: Sun Sep 23, 2007 4:12 pm
Location: Region X
Blog: View Blog (0)

Return to Media and Information Technology

Who is online

Users browsing this forum: No registered users and 1 guest