The Center for Social Data Analytics Colloquium: Riley Tucker

Abstract: Criminologists have long theorized that criminal behavior is driven by visual cues in the environment. Until recently, it has been nearly impossible to construct representative image datasets spanning entire cities.  As such, visual theories of crime have traditionally been tested by sending researchers to small samples of places to manually record observations. Capitalizing on recent advances in AI, this talk introduces a new approach that leverages computer vision to analyze large-scale urban imagery. Using Chicago and Boston as case studies, I show how computer vision AI reveals the visual underpinnings of neighborhood- and address-level crime differences. 

 

Bio: Riley is an Assistant Professor in Criminology and Sociology. His research uses computational methods such as machine learning, natural language processing, and web scraping to better understand cities and the spaces within them. He is particularly passionate about using these methods to test and develop interdisciplinary theories that explain why some urban places experience more crime than others. Prior to joining Penn State, he earned his PhD in Criminology and Justice Policy at Northeastern University and served as a Post Doctoral Fellow at the University of Chicago's Mansueto Institute for Innovation and Data Science Institute. His work has been published in venues such as Journal of Quantitative Criminology, Journal of Research in Crime & Delinquency, and Scientific Data.