Images As Data: Opportunities and Challenges for Automatic Image Analysis

• In Event: Big Data and Machine Learning

Abstract

Social scientists have long argued that images play a crucial role in politics; this role is heightened by the bombardment of images that people experience today. Digitization has both increased the presence of images in daily life and made it easier for scholars to access and collect large quantities of pictures and videos. However, using images as data for social science inference is an arduous task. Political scientists have therefore often turned to other data sources and puzzles, leaving substantive theoretical questions unanswered. Fortunately, recent innovations in computer vision can reduce the costs of using images as data. The role of this paper is twofold. First, we highlight the reasons why images matter in politics and discuss multiple research agendas that would benefit from using images as data. Second, we build on existing computer vision methods to present a set of automatic techniques that will aid political scientists working with images: object detection and recognition, face detection and recognition, and visual sentiment analysis. We use political science examples to illustrate the potential of these applications for the field and provide preliminary open source software that facilitates their implementation.

Authors

• Andreu Casas, University of Washington

• Nora Webb Williams, University of Washington