No RSVP required for current Princeton faculty, staff, and students. Open to members of the public by invitation only. Please contact Jean Butcher at if you are interested in attending a particular lunch.
Politicians and political organizations routinely interact with voters and the public at large using images, yet until recently, computational limitations have precluded efforts to gain systematic knowledge about how images function as a medium of political communication. New developments in machine learning, however, are bringing the systematic study of images within reach. In this paper, we provide a framework for political image analysis with deep neural networks. In addition to introducing neural networks and deep learning methods, we also discuss some of the promises and pitfalls of these techniques for political image analysis moving forward. Using a database of 296,460 photos from the Facebook pages of members of the U.S. House and Senate, we provide two illustrative examples of how these techniques can be used to study how political figures present themselves to the public through images.
Jason Anastasopoulos is CITP’s Microsoft Visiting Professor of Information Technology Policy. Jason’s research focuses on political economy, political institutions, and political methodology. His current research focuses on understanding when, why and how centralized governing bodies delegate policy-making authority. Other research focuses on the interplay between legislative accomplishment and conflicts between interest groups and constituency pressures in the United States.
While at Princeton, Jason plans to study the political economy of cryptocurrencies such as Bitcoin and their potential impacts on political institutions and monetary policy.
Jason’s research interests in political methodology include machine learning methods for text and image analysis with a special interest Bayesian inference for causal inference, deep learning, network analysis and the emerging field of algorithmic game theory. Projects in political methodology include the development of causal inference methodologies for time series data using deep learning methods such as Long Short Term Memory (LSTMs) neural networks and General Adversarial Networks (GANs).