Crime solving and geographic profiling

1
Crime solving and geographic profiling

• J426 Class 27
• April 21, 2009

2
Overview

• Spatial description
• Journey to crime analysis
• Sequential, space-time analysis
• Geographic profiling

3
Nature of the problem, questions to be answered

• Have locations of multiple crimes assumed to be
  committed by same suspect
• Can you predict the location of the next crime
  committed by suspect?
• Can you predict the location of the suspects
  residence?

4
Spatial description

• Basic spatial statistics can be calculated in
  CrimeStat, other programs
• Can indicate estimates of center of set of crime
  locations, possible residence
• Can indicate estimates of extent of set of crime
  locations, area in which more crimes might be
  expected

5
Locations of serial larcenies
From CrimeStat II, Ned Levine Associates
6
Locations of centers of distribution of serial
larcenies
7
Standard deviation ellipse for serial larcenies
8
Animal Movement and Crime Analysis Spatial
Extension

• Animal Movement extension for ArcView 3.x
  developed for wildlife biologists
• Crime Analysis Spatial Extension for ArcGIS 9.0
  implements many of the same analytical
  capabilities

9
Minimum convex polygon home range
From Animal Movement
10
Jennrich-Turner home range
From Animal Movement
11
Spider diagram
From Animal Movement
12
Spider diagram
From Crime Analysis Spatial Extension
13
Journey-to-crime analysis

• Based upon location theory
• Assumes that probability of making trip declines
  with distance
• Given location of serial offenses, construct
  probability surface of likelihood of location of
  residence, given locations of crimes

14
Journey-to-crime travel demand functions in
CrimeStat
From CrimeStat II, Ned Levine Associates
15
Estimating a journey-to-crime travel demand
function
From CrimeStat II, Ned Levine Associates
16
Locations of serial larcenies
From CrimeStat II, Ned Levine Associates
17
Predicted residence using two journey-to-crime
methods
From CrimeStat II, Ned Levine Associates
18
Sequential, space-time analysis

• View and analyze pattern of serial crimes over
  time
• Display movement path
• Calculate movement path statistics
• Predict future movement

19
Create polyline from point file
From Animal Movement
20
Display movement path
From Animal Movement
21
Movement path statistics
From Animal Movement
22
Movement path and statistics
From Crime Analysis Spatial Extension
23
Correlated walk analysis prediction
From CrimeStat II, Ned Levine Associates
24
Correlated walk analysis prediction 2
From CrimeStat II, Ned Levine Associates
25
Geographic profiling

• The development of these methodsespecially
  journey-to-crime analysis, has led to development
  of geographic profiling
• D. Kim Rossmo, Geographic Profiling, 2000

26
Geographic profile

• Includes both quantitative (objective) and
  qualitative (subjective) components
• Objective predictions of offender residence
  location using j-t-c analysis
• Subjective information about hunting style,
  environments of crime sites, physical boundaries,
  etc.

27
Use of geographic profiles

• Subject prioritization
• Linkage with other databases, e.g., registered
  sex offenders, motor vehicle
• Patrol saturation
• Neighborhood canvasses