Sign In

Communications of the ACM

ACM TechNews

Connections Do Count

View as: Print Mobile App Share:
University of Pennsylvania Professor of Computer and Information Science Michael Kearns

The University of Pennsylvania

University of Pennsylvania computer scientist Michael Kearns (pictured) is exploring the connections between social networks and human behavior. Such research could revolutionize how trends and opinions spread through society, says Yahoo!'s Duncan Watts. He says that marketers and much of the public have accepted the idea that society is controlled by a minority of well-connected "influentials" who spread ideas in ways that resemble the spread of a virus. However, Kearns says there is no empirical evidence on how this phenomenon happens.

To learn how influence networks work, Kearns created a social network from a group of 36 subjects, putting each subject at a workstation that was connected to between two and 18 other workstations. Subjects were asked to vote for red or blue, and if they could agree on the same color within one minute everyone would receive a financial reward. Subjects were given different preferences, with some earning more for one color than the other, creating a tension between subjects, with everyone wanting to agree but also wanting them to agree on their color. Kearns found that, depending on the connections and influence, the minority could win over the majority.

Similar research includes a study performed by Watts, which found that human networks are surprisingly unpredictable and quirky, and that people are willing to change preferences to conform to a group. Another study by Cornell University computer scientist Jon Kleinberg focused on how ideas and trends spread throughout populations.

From The Philadelphia Inquirer
View Full Article


No entries found