Features

GPS, Q6: There are many social networking sites, like MySpace, Facebook, LinkedIn, etc that are trying to ride the Web 2.0 wave. What can a researcher like you learn from their success and do they do any interesting link/network analysis?

Kleinberg: Social networking sites are seeking to accomplish many goals at once -- interpersonal, informational, economic, and many others. It will be very interesting to see whether these sites converge to a dominant style of use, and what that will be.

The informational and economic aspects raise many appealing questions. Essentially, when viewed as tools for finding information and getting questions answered, these sites are part of the broader trend toward information-seeking based on the knowledge possessed by people you know, supplementing the much vaster but less verifiable knowledge that resides on the Web at large. In other words, the friends of your friends may not always be able to answer your questions, but when they can, the answer is endowed by this network of friendships with a chain of trust. The trade-offs between this and the more traditional notion of Web search -- and some of the economic consequences of this trade-off -- is a question that Prabhakar Raghavan and I explored in our work on Query Incentive Networks in 2005.

The interpersonal aspects of these sites raise challenging questions as well. If you look at the experience of college undergraduates using Facebook, for example, one of the striking things is the way it really puts the idea of a social network on center stage. For many generations, people have moved away from home to places like college campuses and been implicitly aware of the ways in which they were forming connections, embedding themselves in a social structure. But the underlying social network itself was always latent, invisible, and to some extent, unknowable. Now, to have constant access to Facebook, to have your place in the social network depicted so explicitly on a computer screen, updating itself as you form these connections, following your progress like a kind of scoreboard -- this foregrounding of the social network is really a new phenomenon. It will be interesting to see where the feedback effects of such an explicitly visualized network will take us.

What's this?