Slide 1
Slide 1 text
Marco De Nadai, Yanyan Xu, Emmanuel Letouzé,
Marta C. González, Bruno Lepri
Socio-economic, environmental, and
mobility conditions associated with crime:
A study of multiple cities
ICCS 2018
Slide 2
Slide 2 text
2
Urban population is exploding
Gomez-Lievano et al. "Explaining the prevalence, scaling and
variance of urban phenomena." Nature Human Behaviour (2017)
Slide 3
Slide 3 text
3
How can we describe crime?
Socio-economic
characteristics
Built environment Mobility
Slide 4
Slide 4 text
Socio-economic
characteristics
Slide 5
Slide 5 text
The social-disorganization theory
Crime is a result of lack of cooperation and trust
• Economic deprivation
• Ethnic heterogeneity
• Residential instability
5
Sampson, R. J., et al. (1989). Community structure and crime: Testing social-
disorganization theory. American journal of sociology
SOCIAL
DISORGANIZATION
Slide 6
Slide 6 text
Bayesian Poisson model
6
Leroux et al. "Estimation of disease rates in small areas: a new mixed model for
spatial dependence." Statistical models in epidemiology, the environment, and
clinical trials.
log &
= ( *
*
,
*-.
+ CAR process
Auto-correlation matrix
Features
(e.g. disadvantage)
Crime
in a district
(ground truth)
SOCIAL
DISORGANIZATION
Slide 7
Slide 7 text
Social disorganization - Results
7
SOCIAL
DISADVANTAGE
ETHNIC
MIX
RESIDENTIAL
STABILITY
THEORY
THEORY
THEORY
Slide 8
Slide 8 text
Social disorganization - Results
8
SOCIAL
DISADVANTAGE
ETHNIC
MIX
RESIDENTIAL
STABILITY
THEORY
THEORY
THEORY
Slide 9
Slide 9 text
Social disorganization - Results
9
SOCIAL
DISADVANTAGE
ETHNIC
MIX
RESIDENTIAL
STABILITY
THEORY
THEORY
THEORY
Slide 10
Slide 10 text
Social disorganization - Results
10
SOCIAL
DISADVANTAGE
ETHNIC
MIX
RESIDENTIAL
STABILITY
THEORY
THEORY
THEORY
Slide 11
Slide 11 text
Social disorganization - Results
11
THE SOCIAL DISORGANIZATION
IS NOT ENOUGH
Slide 12
Slide 12 text
The built environment
Slide 13
Slide 13 text
The built environment
Jane Jacobs has been one of the most influential
people in urban planning
13
Jacobs, Jane. The death and life of great American cities. Vintage, 1961
BUILT
ENVIRONMENT
Slide 14
Slide 14 text
The built environment
Jane Jacobs has been one of the most influential
people in urban planning
14
Jacobs, Jane. The death and life of great American cities. Vintage, 1961
Informal surveillance: Guardianship by
ordinary citizen, and police
BUILT
ENVIRONMENT
Slide 15
Slide 15 text
The built environment
Four essential conditions:
1. Land use mix
15
Land use mix
Jacobs, Jane. The death and life of great American cities. Vintage, 1961
BUILT
ENVIRONMENT
Slide 16
Slide 16 text
The built environment
Four essential conditions:
1. Land use mix
2. Small blocks
16
Small blocks
Jacobs, Jane. The death and life of great American cities. Vintage, 1961
BUILT
ENVIRONMENT
Slide 17
Slide 17 text
The built environment
Four essential conditions:
1. Land use mix
2. Small blocks
3. Aged buildings
17
Aged buildings
Jacobs, Jane. The death and life of great American cities. Vintage, 1961
BUILT
ENVIRONMENT
Slide 18
Slide 18 text
The built environment
Four essential conditions:
1. Land use mix
2. Small blocks
3. Aged buildings
4. Density
18
Density
Jacobs, Jane. The death and life of great American cities. Vintage, 1961
BUILT
ENVIRONMENT
Slide 19
Slide 19 text
The built environment
Four essential conditions:
1. Land use mix
2. Small blocks
3. Aged buildings
4. Density
19
De Nadai, Marco, et al. "The death and life of great Italian cities: a mobile phone
data perspective." WWW, 2016.
OpenStreetMap + census!
BUILT
ENVIRONMENT
Slide 20
Slide 20 text
Bayesian Poisson model
20
log &
= ( *
*
,
*-.
+
CAR process
Spatial matrix
Features
(e.g. land use mix)
Crime
in a district
(ground truth)
De Nadai, Marco, et al. "The death and life of great Italian cities: a mobile phone
data perspective." WWW, 2016.
Slide 21
Slide 21 text
21
Built environment – Results (errors)
Model Bogota Boston Los Angeles
Social disorganization 163 374 356
Built environment 109 352 317
Slide 22
Slide 22 text
22
Built environment – Results (errors)
Model Bogota Boston Los Angeles
Social disorganization 163 374 356
Built environment
(+ Social disorganization)
109
(104)
352
(205)
317
(333)
Slide 23
Slide 23 text
23
Built environment – Results (errors)
Model Bogota Boston Los Angeles
Social disorganization 163 374 356
Built environment
(+ Social disorganization)
109
(104)
352
(205)
317
(333)
CRIME ESTIMATION
FROM URBAN DATA
Slide 24
Slide 24 text
24
Built environment - Discrepancies
Small blocks
Just an example…
Slide 25
Slide 25 text
25
Built environment - Discrepancies
THE PLACE IS NOT ENOUGH
Just an example…
Slide 26
Slide 26 text
Mobility
Slide 27
Slide 27 text
Mobility
27
Graif, Corina et al. "Neighborhood isolation in Chicago: Violent crime effects on
structural isolation and homophily in inter-neighborhood commuting networks."
Social networks 51 (2017)
• People are everyday exposed to different
conditions and places
• Poorly connected neighborhoods weaken
cooperation
• Mobility affects crime
MOBILITY
Slide 28
Slide 28 text
Mobility
28
Digitalization of our life => mobile phone data
Jiang, Shan, et al. "The TimeGeo modeling framework for urban mobility without
travel surveys." Proceedings of the National Academy of Sciences 113.37 (2016)
MOBILITY
Slide 29
Slide 29 text
Bayesian Poisson model
29
O/D
NETWORK
C B
A
34
21
90
MOBILITY
Slide 30
Slide 30 text
Bayesian Poisson model
30
LOW
HIGH
POPULATION
O/D
NETWORK
C B
A
34
21
90
MOBILITY
Slide 31
Slide 31 text
Bayesian Poisson model
31
LOW
HIGH
POPULATION
O/D
NETWORK
log &
= ( *
*
,
*-.
+
Crime
in a district
(ground truth)
C B
A
34
21
90
MOBILITY
Slide 32
Slide 32 text
Results (errors)
32
Model Bogota Boston Los Angeles
Social disorganization 163 374 356
Built environment 109 352 317
Mobility 141 310 280
Slide 33
Slide 33 text
Results (errors)
33
Model Bogota Boston Los Angeles
Social disorganization 163 374 356
Built environment 109 352 317
Mobility 141 310 280
Full model 82 267 240
PLACE + PEOPLE + DYNAMISM
Slide 34
Slide 34 text
34
How can we describe crime?
Socio-economic
characteristics
Built environment Mobility
Slide 35
Slide 35 text
35
How can we describe crime?
Socio-economic
characteristics
Built environment Mobility
✓ Precise data ✓ Availability
✓ Unbiased
(population)
✗ Rarely updated
✗ Availability
✗ Bias over volunteers ✗ Availability
Slide 36
Slide 36 text
36
How can we describe crime?
Socio-economic
characteristics
Built environment Mobility
ALL BLOCKS TOGETHER
BETTER DESCRIBE CRIME
THEORY HAS DISCREPANCIES
OVER DIFFERENT CITIES
Slide 37
Slide 37 text
37
How can we describe crime?
New theory
Slide 38
Slide 38 text
Feedback time!
@denadai2