Abstract: Dr.Nathan Hodas '04

Using Social Network Data to Understand Human Behavior

As a growing fraction of our lives is conducted online, we create ever richer trails of digital breadcrumbs for learning about statistical human behavior.  By analyzing different online social networks and citation patterns in papers using the techniques of statistical physics, we have concluded that our fundamental cognitive limitations constrain the spread of viral information.  Each person has a finite amount of energy they are willing or capable of devoting to decision making each day, just as we are only capable of lifting a finite amount of weight in day.  We have evolved to be “cognitive misers.”  As a result, the ease of discovery of information is the dominant factor in facilitating information dissemination.  I will discuss these result and how to apply these lessons to forecast decision making online.