Analyzing Late-Night Risk Behaviors of Women Using Graph Analysis: A Reexamination of Crime Pattern Theory

• In Event: Poster Session 1

Abstract

In Japan, late-night behaviors such as walking more than 1 km from a station to home and stopping at a convenience store on the way home are cited as risk factors for sexual crime victimization. We analyze the extent to which women actually engage in risky behaviors late at night using human flow data from urban and suburban areas. In the analysis, the trajectory of human movement is represented as a graph with workplaces, stations, and homes as nodes and travel routes as links, and is quantified using a graph analysis approach. This allows us to quantitatively understand the movements of women during late-night hours not as "points" but as “lines”. Additionally, we confirm the overlap between areas with high rates of sexual victimization of women and areas where theoretically risky behaviors are concentrated. We also examine which types of actual movement trajectories can be more dangerous. The results are discussed in accordance with crime pattern theory.

Author

• Tomoya Ohyama, University of Tokyo