Explaining Crime Diversity with Google Street View

Crime diversity is a measure of the variety of criminal offenses in a local environment, similar to ecological diversity. While crime diversity distributions have been explained via neutral models, to date the environmental and social mechanisms behind crime diversity have not been investigated. Bui...

Descripción completa

Guardado en:  
Detalles Bibliográficos
Autor principal: Khorshidi, Samira (Autor)
Otros Autores: Tita, George E. ; Carter, Jeremy ; Mohler, George
Tipo de documento: Electrónico Artículo
Lenguaje:Inglés
Publicado: 2021
En: Journal of quantitative criminology
Año: 2021, Volumen: 37, Número: 2, Páginas: 361-391
Acceso en línea: Volltext (lizenzpflichtig)
Journals Online & Print:
Gargar...
Verificar disponibilidad: HBZ Gateway
Palabras clave:
Computervision
Crime diversity
Google Street View
Geographically weighted regression
Computer vision
Descripción
Sumario:Crime diversity is a measure of the variety of criminal offenses in a local environment, similar to ecological diversity. While crime diversity distributions have been explained via neutral models, to date the environmental and social mechanisms behind crime diversity have not been investigated. Building on recent work demonstrating that crime rates can be inferred from street level imagery with neural network computer vision models, in this paper we consider the task of inferring crime diversity through street level imagery.
ISSN:1573-7799
DOI:10.1007/s10940-021-09500-1