Frontiers in Spatial and Spatiotemporal Crime Analytics—An Editorial (original) (raw)

2017, ISPRS International Journal of Geo-Information

Environmental criminological theory is well-developed [1,2] but analytical techniques to explore and model crime incidents are lagging behind. Due to the emergence and accumulation of a wide range of environmental data [3], volunteered geographic information [4], unstructured textual information [5], and (big) statistical data [6], among others, it is of particular relevance to keep pace with these developments and make use of such spatial and temporal data. The integration of such data is highly relevant for space-time crime analytics, potentially providing new insight into local crime hotspots supporting law enforcements to combat crime. While geographic information system-based methods [7] and spatial statistical approaches have nowadays gained momentum to map crime patterns [8], advanced data-driven computational methods (e.g., machine learning, Bayesian spatiotemporal models) are still in their infancy and are far from being mainstream (e.g., [9,10]). However, other disciplines provide evidence that these approaches are highly capable of solving classification problems, conducting inference, forecasting, and extracting patterns hidden in data that are otherwise overseen by basic methods [11,12]. Therefore, the amalgamation of criminology with computational methods seems to be a rational next step in the research agenda. To address this, the prime aim of this Special Issue Frontiers in Spatial and Spatiotemporal Crime Analytics was to publish original research or review papers in order to stimulate further discussion on the development and application of the latest data-driven scientific advances to understand crime patterns and criminal behavior, their dynamics over time and across space, and the underlying key mechanisms. We anticipate that this methodological progress will yield more reliable risk assessments and more accurate predictions of crime as demanded by criminal justice agencies and needed for evidence-based criminal justice decision-making. The collection of paper provides a selection of actual approaches useful for, but not limited to, audiences that include researchers, postgraduates, and professionals. This Special Issue is a follow-up publication on a Special Issue [13] on crime mapping principles and we believe that this new collection of papers will contribute to the contemporary research agenda on crime modeling from a computational and data-driven perspective. As indicated by the response to the call for manuscripts, it appears that the intention of this Special Issue received wide approval. By the end of December 2016 (submission deadline), a total of ten manuscripts were submitted. Each manuscript was evaluated through a single-blind review process by at least two international experts. For the review process, the standard MDPI review guidelines were used. Besides the innovative aspect of the research, the scientific quality of the research weighted heavily on the decision of whether or not a manuscript was accepted or rejected. In cases where major revisions were requested by reviewers and to guarantee high scientific quality, a second round of review of the revised manuscript by at least one of the original reviewers or an alternative reviewer was conducted. If reviews called for minor revisions, then a second round of reviews was not done. Instead,