Detecting spatial-temporal clusters of violent behavior in South Korea with space-time permutation scan statistics
| Autor: | Yunho Yeom |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
030505 public health
Public Administration Injury control Computer science Space time Law enforcement Poison control 02 engineering and technology Pathology and Forensic Medicine 03 medical and health sciences Permutation Identification (information) 020204 information systems Georeference Statistics 0202 electrical engineering electronic engineering information engineering 0305 other medical science Cluster analysis Law |
| Zdroj: | Policing: An International Journal. 42:490-502 |
| ISSN: | 1363-951X |
| DOI: | 10.1108/pijpsm-07-2018-0085 |
| Popis: | Purpose The purpose of this paper is to detect spatial-temporal clusters of violence in Gwanak-gu, Seoul with space-time permutation scan statistics (STPSS) and identifies the temporal threshold for such detection to alert law enforcement officers quickly. Design/methodology/approach The case study was the Gwanak Police Station Call Database 2017 where civilian calls reporting violence were georeferenced with coordinated points. In analyzing the database, this study used the STPSS requiring only individual case data, such as time and location, to detect clusters of investigated phenomena. This study executed a series of experiments using different minimum and maximum temporal thresholds in detecting clusters of violence. Findings Results of the STPSS analyses with different temporal thresholds detected spatial-temporal clusters in Gwanak-gu. Number, location and duration of clusters depended on the temporal settings of the scanning window. Among four models, a model allowing the possible clusters to be detected within a 7-day minimum and 30-day maximum temporal threshold was more representative of reality than other models. Originality/value This study illustrates the clustering of violence with the STPSS by detecting spatial-temporal clusters of violence and identifying the appropriate temporal threshold in detecting such clusters. Identification of such a threshold is useful to alert law enforcement officers quickly and enables them to allocate their resources optimally. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|