Social media sites like Twitter enable users to engage in the spread of contagious phenomena: everything from information and rumors to social movements and virally marketed products. The dynamics of these phenomena have been studied extensively from a theoretical perspective, but a gap exists between models and empirical studies of social contagions. In this talk, I will present work towards bridging that gap by first describing the complex contagion model for the propagation of viral phenomena; then, demonstrating a critical mass effect emergent from this model that may explain the fragile dynamics of empirical contagions; and finally, using the model to analyze the dynamics of real-world contagions including tags on the Flickr photo sharing network and political hashtags used by Russians on Twitter. I end with some thoughts about the challenge of capturing the complex dynamics of empirical contagions and future modeling work towards meeting this challenge.
Bio: Vladimir Barash is a Senior Researcher and Engineer at Morningside Analytics. He has received his Ph.D. from Cornell University, where he studied with Michael Macy, Jon Kleinberg, and Claire Cardie. Both at Cornell and now at Morningside, Vladimir’s research interests focus mainly on the intersection of social media and large-scale social phenomena, such as online political activism in Russia or the cross-cultural patterns of emoticon use in Twitter. Vladimir’s work has received awards at the International Conference for Weblogs in Social Media and Bits on Our Minds. In his spare time, Vladimir reads and writes fiction, and blogs at: vlad43210.wordpress.com.