U tackles social network intricacies

A computer science team’s work could also help identify terrorist rings.

by Jennifer Bissell

While social media is often used haphazardly, one University of Minnesota computer science team has developed a model that takes a closer look at the intricacies and relationships it shows.

The research has gotten the attention of interested entities, both traditional (businesses looking for a leg up in marketing) and brand new (the U.S. government, in the hopes of identifying terrorist rings).

Typically, social networks like Facebook, Wikipedia and Foursquare show very static relationships between users, but with professor Shashi ShekharâÄôs model, more dynamic relationships can be revealed.

“Facebook is a trivial example of social networks,” Shekhar said. “On Facebook, either youâÄôre a friend or youâÄôre not. There are no shades of friendship.”

But with ShekharâÄôs model, which measures relationships over time, questions about new leadership, changes in trust and the cohesiveness of social groups could be answered.

Dev Oliver, a graduate student working on the project, explained further that massively multiplayer online role-playing games with multiple levels of friendship âÄî visitor, friend or trustee âÄî are a good example of where more complicated data could be synthesized by their new model.

“The thing about these social networks is that they are very big,” Oliver said. “So we need computational techniques in order to handle the sheer size of the network and answer meaningful questions.”

Additionally, building this kind of model could have profound effects on businessesâÄô marketing strategies, as it would be easier to see the most influential leaders in different locations.

Social media and digital strategist Nicole Harrison, who helps businesses build social media platforms, said this kind of knowledge could be very helpful.

“Social media is about having a strategy and knowing what youâÄôre doing to connect with the network,” Harrison said. “The more valuable you are, the more theyâÄôll listen to you.”

Thus, knowing the social patterns of how people latch on to different ideas and people, which ShekharâÄôs model can track, could be essential to a new wave of marketing.

“[Social media] is not going to go away but get more embedded into what we do,” Harrison said. “It will revolutionize the way we do things. You can get people to come together and help each other out.”

But businesses are not the only entities interested in the analysis of social networks. The U.S. departments of Defense, Justice and Transportation have taken an interest in Shekhar and OliverâÄôs project as well.

Shekhar said the departmentsâÄô interests varied from planning emergency evacuation systems âÄî where users would be able to upload information about blocked roads and safe havens âÄî to mapping “human terrain,” such as criminal and terrorist networks.

“From this point forward they want not only physical maps,” Shekhar said, “but they also want to understand the social networks” maps that show, for example, “here are the tribes, and this tribe here is the decision makers.”

Counterterrorism professor David Knoke said heâÄôs glad to hear this kind of modeling is being researched but cautioned the data required to make a model mapping terroristsâÄô work would be hard to obtain.

“This is the kind of basic research the counterterrorism organizations should be engaging in,” Knoke said. “But terrorists are not going to volunteer this kind of information [the model will need].”