"When studying different behaviors in online social networks, such as flow of information, bargaining power, flow of influence, it is useful to be able to characterize a node with respect to the whole network," using certain measures that can "describe various aspects of the 'importance' of a node in a network: how connected the node is, how easily it can reach other nodes, how much it mediates the connection between other nodes, etc.," he added.
The researchers also are studying various privacy aspects of online social networks, and how its members search and acquire new knowledge based on their personal preferences, "meeting" other members with similar interests. For the latter research, the scientists looked at the operation of Last.fm, a music network that creates user profiles based on listening habits.
"The profiles are used by Last.fm to create lists of possible new friends, and possible artists, based on compatibility between the profiles," Stefanescu said. "You want to help users expand their musical horizons, and make new friends. As it turns out, the success is somewhat variable; the compatibility doesn't always work out very well."
Finally, the scientists are studying interrelationships between offline characteristics of users, such as gender or culture, and their internet preferences. They want to better understand the correlation between these offline characteristics and their online behavior, such as communication patterns and relationship building.
"You can invest in many different scenarios by exploring social networks," Stefanescu said.
Follow U.S. News Science on Twitter.