Abstract

In this study, we aimed to investigate the effectiveness of big data-driven predictive policing, one of the latest forms of technologybased policing, and also the risks of data concentration on police forces or algorithmic bias. In order to properly weigh the benefits and risks, we first conducted a systematic review of the effectiveness of big data-driven predictive policing, based on the terminology of the research topic, and finally extracted 161 articles. We classified them into four categories based on the strength of the evidence and examined the evidence of effectiveness in each category. We concluded that while it is encouraging to see a range of studies, given the significant concerns about big data-driven predictive policing, evidence of effectiveness that can be used by policymakers still needs to be supported by more research, as only 6 out of 161 were evidencestrong in our study categorisation.