Studying Network Structural Changes Using Information Event Signatures

Abstract

The diffusion of information is important: It impacts commerce. It influences government. It connects people in new ways. The distributed nature of our digital social networks means that traditional gatekeepers (newspapers, radio, television, and governments) lose some control over the flow of information, while new gatekeepers emerge quickly in networks of individuals who share interests or grievances. Using exploratory data analysis and confirmatory statistics to analyze over 64 million Occupy movement tweets, this dissertation makes four essential contributions that enhance our understanding of the relationship between the flow of information and the dynamics of social networks. First, based on a large set of Twitter data related to the Occupy Wall Street movement, it introduces a parameterized signature model of individual information flows. Second, it demonstrates that both the path of the information flow and the changes in the structure of the network, as measured by the growth of network gatekeepers within the Occupy movement, are related to parameters of the model. Third, the analysis suggests that the Occupy gatekeepers recursively extend their reach by repeatedly promoting information that users shared deep into Twitter's social network of followers. Fourth, the model provides the initial steps towards a theory explaining the process by which social network dynamics and information flows interact. This model's capacity to identify information flows deep in networks and to predict trends that potentially alter those networks will prove useful to individuals, to organizations, and to governments.