Area of Study: I develop machine learning methods for relational datasets drawn from social networks. My focus is on distinguishing the social phenomenon of homophily and social influence. Homophily is the tendency for individuals who share certain traits to develop relationships in the network (links), while social influence is the tendency for individuals to adopt traits from their neighbors in the network. While these effects represent very different processes in the evolution of social networks, they both yield networks with similar properties. I have developed a randomization test that given several time slices of a network can distinguish whether homophily and/or social influence are occurring in the network evolution.
