다중 행위자 기반 학교주변 가로의 범죄위험 지점 탐색모델 개발에 관한 연구

Abstract

The purpose of this study was to develop a multi-agent-based model searching crime risk spots in school zone streets. The specific location data with crime risk have been were hardly, as meso-scale or local space, identified within school zone streets. The review of precedent studies on school zone crimes, situational crime theories and agent-based modeling and simulation approaches related to crime situations suggested directions of the development. Overall process included four parts; data input, modeling, simulation and result recording. Two networks representing walkways and vehicular roads considering agents’ travel methods and the use types of facilities configured a combined spatial network model. GIS system linked to the model enabled it to derive data for spatial networks and the facility use types. Students, attackers, school staffs, police officers driving patrol cars, police stations and pedestrians were implemented as behavioral agents composing crime situations. Finally, a spatio-temporal simulation of the proposed model searching crime risk spots in school zone streets was able to specify the spots with potential crime situations. This study also applied the implemented model and the simulation procedure to an existing school zone conditions and found crime risk spots frequently identified from a field survey on the area.