Many scientific disciplines have benefitted from the availability of big datasets to develop algorithm supported solutions. Recently, this trend has penetrated the fields of crime and police research. The presentation highlights use cases of big data computation and HPC for typical datasets in crime science: crime records, emergency call data, and police GPS data. The focus lies on spatiotemporal applications (i.e., geocoding, map matching, spatial and temporal algorithms). The datasets come from a collaborative project with a Belgian police force and encompass approximately 200,000 crime records and 400 million individual GPS datapoints (x and y coordinates + timestamp). The project aims at establishing the crime preventive effect of police patrols through a longitudinal research design. Besides offering novel computational solutions for crime scientist, these kinds of datasets introduce important ethical considerations as well as potential biases in human-led data entry and collection. These considerations underline the need for social scientists to become more literate in the computational sciences and support the framework of ‘crime science’ as a discipline of exact methods and data sources.
Many scientific disciplines have benefitted from the availability of big datasets to develop algorithm supported solutions. Recently, this trend has penetrated the fields of crime and police research. The presentation highlights use cases of big data computation and HPC for typical datasets in crime science: crime records, emergency call data, and police GPS data. The focus lies on spatiotemporal applications (i.e., geocoding, map matching, spatial and temporal algorithms). The datasets come from a collaborative project with a Belgian police force and encompass approximately 200,000 crime records and 400 million individual GPS datapoints (x and y coordinates + timestamp). The project aims at establishing the crime preventive effect of police patrols through a longitudinal research design. Besides offering novel computational solutions for crime scientist, these kinds of datasets introduce important ethical considerations as well as potential biases in human-led data entry and collection. These considerations underline the need for social scientists to become more literate in the computational sciences and support the framework of ‘crime science’ as a discipline of exact methods and data sources.
Speakers: Philipp M. Dau