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...

Full description

Saved in:  
Bibliographic Details
Authors: Khorshidi, Samira (Author) ; Tita, George E. (Author) ; Carter, Jeremy (Author) ; Mohler, George (Author)
Format: Electronic Article
Language:English
Published: 2021
In: Journal of quantitative criminology
Year: 2021, Volume: 37, Issue: 2, Pages: 361-391
Online Access: Volltext (lizenzpflichtig)
Journals Online & Print:
Drawer...
Check availability: HBZ Gateway
Keywords:
Computervision
Crime diversity
Google Street View
Geographically weighted regression
Computer vision
Description
Summary: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