Uncovering crime patterns using location data

The way people move around in cities gives us clues as to when many crimes are committed and in which hotspots, as ETH researchers have demonstrated using anonymised data from  location technology plattforms. When and where does crime arise in cities? To answer this question, criminologists have previously relied on rather static models. Crime has been linked, for example, to the structure of the resident population or to the use of land in a neighbourhood. The influence that mobility has on the incidence of crime was previously an unknown quantity. Now, however, researchers from ETH Zurich, the University of Cambridge and New York University have been able to demonstrate for the first time that crime is directly related to how many people are in a city, where they are and where they are going. Cristina Kadar, a computer scientist and former doctoral student at the Mobiliar Lab for Analytics at ETH, led the study. She recently presented the results at a (virtual) conference on computational social science.