Year: 2022
Type: dissertation
Abstract: How can we learn about the nature of relationships that people form online? We argue that topic models can produce content-based representations of social relationships that are human-interpretable and useful for computational analysis. Given communication data (Twitter, Facebook, e-mail), we attempt to detect social influence at two levels of aggregation. At the micro level, we explore the connection between influence and Granger-causality. At the meso level, testing for influence is reframed as predicting a person's behavior using information from the social environment.
Language: English
Subfield: Statistical and Nonlinear Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7071718220
