University of Glamorgan
The Dynamic Spatial Disaggregation Approach to Geo-Temporal Crime Forecasting (Friday: 10:00 - 11:30 a.m.)
Answers to the questions “how many,” “where,” and “why” in crime prediction applications represent important elements of crime
prevention operations. Two novel geo-temporal crime modeling and forecasting approaches are proposed: the hierarchical profiling
approach (HPA) and the dynamic spatial disaggregation approach (DSDA). HPA explicitly models the influences of societal events
on crime levels within the forecasting framework. DSDA employs clustering methods to disaggregate the forecasts dynamically,
enabling the influences of weekdays and exogenous variables on the spatial distribution to be modeled. Results using HPA and
DSDA on multiyear daily criminal damage incidences in the City of Cardiff, UK, are presented.
Simon Fraser University
The Ambient Population and Crime Analysis (Friday: 3:30 - 5:00 p.m.)
The proliferation of GIS/science in conjunction with the availability of spatially referenced crime data has made the spatial
analysis of crime advance significantly over the past decade. Most often, these analyses involve the plotting of crime points
on a map or aggregating the crime points within census boundary units for analyses with census variables. In this paper, the
ambient population (a 24-hour average estimate of the population present in a spatial unit) is used in conjunction with spatially
referenced crime data and census variables to gain new and significant insight into the spatial dimension of crime.
University College London
Neighborhood Crime and Antisocial Behavior—A National Pilot in the United Kingdom (Saturday: 8:00 - 9:30 a.m.)
This presentation will draw upon findings of a national study commissioned by the Audit Commission to research the use of
spatial analysis and geo-demographics in England and Wales. Ten case-study areas were selected in five diverse regional partnerships.
The research is founded on a U.K. policy drive toward neighborhood policing and explores the translation of geo-demographic
research from business and academia into the public sector. Innovative methods of analyzing small-area demographics for policing
will be presented, prior to reflection upon the lessons learned from both the national study and feedback from a tour of regional
showcase events to practitioners.
Department of Statistics - Universidade Federal de Minas Gerais
Space-Time Crime Clusters Identification (Thursday: 1:30 - 3:00 p.m.)
We propose a new type of scan statistic to identify and evaluate the statistical significance of localized clusters in space
and time, rather than just testing for the presence of clustering, such as in Knox-Mantel test of CRIMESTAT. The method can
be used either retrospectively to understand crime patterns or prospectively to detect emerging clusters. The method has been
implemented in free available software. We illustrate our method and the program with space-time data of different types of
robberies in Belo Horizonte, a Brazilian city.
DWI Resource Center, Inc.
Data Visualization–Based Inference of Time-Based Drunk Driver–Preferred Driving Routes (Thursday: 3:30 - 5:00 p.m.)
DWIRC has used data visualization techniques and geo-coding of drunk driver crashes, drunk driver residencies, and alcohol
establishment densities to create a map of “preferred” drunk driving routes in Bernalillo County, New Mexico. The ability
to show both crash frequency and alcohol involved to total crash ratios, coupled with drunk driver and alcohol establishment
density can assist in optimizing limited law enforcement resources through strategic placement of DWI checkpoints and increased
saturation patrols in specific areas. In addition, because of the amount of data involved, the mapping of alcohol establishments,
alcohol-involved accident ratios, and drunk driver residencies allows citizens an efficient method of choosing those routes,
areas, and times considered more dangerous and thus avoided.
Observatoire National de la Délinquance
National Criminal Statistics in France: The Geographic Way to Share the Data (Friday: 3:30 - 5:00 p.m.)
France is very late in the crime mapping field. However, the geographic approach was born at home, in France (Quételet, Guerry)
–and England, Yes (Mayhew) it’s true. Currently, disseminate/produce sincere criminal statistics which is state of the art.
In this goal, there is not any doubt the internet map is the best answer. This presentation will show the French statistic
method to analyse 107 index of crimes across the French uniform crime report: the “etat 4001” statistic. The OND has built
three indicators and many categories and subcategories of indicators to understand criminal phenomena. Then we plan the internet
project to spread analyse, information and data with the map interface: the Carto.Net project. Finally, this presentation
will approach the future studies, notably “criminals in Paris, origins and destinations”.
Avinash Singh Bhati
Justice Policy Center, The Urban Institute
Multi-Moment Spatial Analysis of Violence (Friday: 1:30 - 3:00 p.m.)
Scholars studying violence typically model only the first moment of the dependent variable—its expected value. This is an
unfortunate limitation because other moments of violence can contain important information about the phenomenon. Of particular
interest is whether and to what extent there exists spatial autocorrelation among multiple moments. This presentation will
demonstrate the application of a semi-parametric approach for investigating multi-moment spatial dynamics of the homicide
rate. The approach offers clearer insights into these dynamics and allows researchers to test various theoretically motivated
hypotheses about the process under study.
Massachusetts State Police
Lies, Damned Lies, and Mapping NIBRS Data (Thursday: 1:30 - 3:00 p.m.)
The National Incident-Based Reporting System (NIBRS) provides a rich data set for analysis but without specific geographic
identifiers. The addition of incident address information permits mapping of crime data, and these data can be used for tactical
purposes. However, important issues of data quality and accuracy must be addressed before this data set can be productively
used. This presentation will give an example of the tactical use of NIBRS data and also examine several issues concerning
data quality in a shared data environment.
Ohio Department of Rehabilitation and Correction
A GIS Approach to the Assignment of Supervision Cases in Franklin County, Ohio (Friday: 3:30 - 5:00 p.m.)
Analysis of address data for offenders under Ohio Adult Parole Authority supervision reveal substantial geographic dispersion
within officer caseloads, complicating time management and potentially undermining more effective supervision strategies.
Previous geographically based systems failed due to offender moves and districts focused on bad neighborhoods. This proposal
divides Franklin County into four wedge-shaped districts of zip codes. Addresses are geo-coded, and analysis by supervision
level shows comparable proportions of higher risk offenders across the proposed districts. Analysis on offender address changes
suggests that most movements can be accommodated through small buffer zones, minimizing reassignment of cases.
Vector Research Group, LLC
A New Approach to Geographic Profiling (Friday: 10:00 - 11:30 a.m.)
Current geographic profiling models use predefined equations or algorithms to predict offender home location. These models
are limited because they do not effectively incorporate important environmental conditions into the prediction algorithm.
Addition of environmental conditions to the geographic profiling model results in a more accurate analysis requiring less
data than needed using current methodology. In some cases, only two or three sites are needed to obtain an accurate geographic
profile. This presentation covers the details of the new methodology, and specific case studies will be presented.
Targeted Law Enforcement in Camden, New Jersey: Diffusion of Benefits and Displacement of Crime (Friday: 3:30 - 5:00 p.m.)
In 2003, Johnson and Bowers presented a weighted displacement model for measuring the impact, spatial displacement, and diffusion
of benefits resulting from crime prevention initiatives. Although useful for exploring large-scale studies, we adapt and simplify
their approach to examine a small-scale saturation patrol operation in Camden, New Jersey. This paper examines target, displacement,
and control area trends in violent crime, gun crime, drug crime, burglary, and vehicle crime. We will present how to use this
approach to evaluate a policing operation and will make available a Microsoft Excel spreadsheet to calculate the Johnson and
Bowers weighted displacement quotients.
West Midlands Police
The Self-Containment of Crime—Area-Based Analysis of Journey-to-Crime Origin and Destination Trips and Implications for Police
and Partnership Work (Thursday: 1:30 - 3:00 p.m.)
Self-containment is an area-based index that measures the extent to which crime is committed locally within an area. It is
simple to calculate, and when analyzed in conjunction with other police, geo-demographic, lifestyle, and socioeconomic data
sets, it can prove a useful indicator to guide police and other partnership agency activity. This paper will examine how self-containment
is calculated, how it varies across different communities, and how it can be used by police and other agencies at a strategic
and local level.
Mapping the Cost of Crime—Area-Based Analysis of the Cost of Crime and Implications for Police and Partnership Work (Saturday:
8:00 - 9:30 a.m.)
This presentation will build on recent research material on the cost of crime (e.g., Dubourg and Hamed, 2005). West Midlands
Police have taken the financial costs indicated in this research and mapped it at local levels to see how placing a financial
weight on crimes can help agencies assess the “harm” crime causes local neighborhoods and communities. By attempting to map
“harm,” police managers can be better informed to prioritize strategic and local activity, deploy resources accordingly, and
more importantly, assist in identifying areas for partnerships.
Frances Frick Burden
Pennsylvania State University
Reexamining Social Disorganization: An Analysis of Recidivism Patterns of Parolees in Georgia (Friday: 3:30 - 5:00 p.m.)
Researchers from the social disorganization school have shown that neighborhoods with high levels of poverty, ethnic/racial
heterogeneity, and residential mobility also suffer from high crime rates. This study examines the effects of socially disorganized
neighborhoods on parole recidivism and attempts to answers two question: (1) Do the traditional measures of social disorganization
predict individual parole recidivism? (2) Are similar variables and new techniques (diffusion and distance-based variables)
better measures of neighborhood social disorganization than current standard variables? Preliminary findings suggest that
while traditional social disorganization measures can predict recidivism, the new variables are stronger predictors.
Winston-Salem State University Center for Community Safety
Community Safety Mapping (Friday: 1:30 - 3:00 p.m.)
Since March 2005, the Transforming Communities Research Laboratory has providing basic GIS training and data acquisition to
the citizens of Winston-Salem, North Carolina, and the surrounding area through a unique “storefront” GIS lab in the heart
of Winston-Salem’s downtown renaissance and renewal. More than 250 people have come in for a free workshop on GIS and community
research. Then they come back to the lab and begin looking at our data sets about their own communities. These sets include
crime data provided by police so residents can look at and/or create their own crime maps. For an in-depth look at what is
going on, they can call on the staff at the Center for Community Safety and involve students and staff to get a better understanding
of what their neighborhoods look like and, more importantly, what they can do to help. This presentation will illustrate the
concept of getting data into the community and present several different project examples from community participants.
University College London, Jill Dando Institute of Crime Science
Searching for Effect: An Exploration of the Local Spatial-Temporal and Proportionate Demographic Use of Stop and Search and
Its Relationship With Crime (Friday: 1:30 - 3:00 p.m.)
Stop and search plays an active role in policing. Considering that stops and searches are most typically used in a non-random
manner, little is known about the profile and productivity of their localized use, how they relate to patterns of crime,and
whether they are proportionately applied when local demographic factors are considered. This paper presents findings from
a major U.K. research study that explores whether searches are geographically and temporally concentrated, the effect that
geographical and temporal targeting of search activity has on its effectiveness, and whether the targeting of searches is
applied proportionately across demographic groups.
The Appliance of Science: Spiriting Crime Mapping with Crime Science (Friday: 3:30 - 5:00 p.m.)
Crime science is about applying a methodical, hypothesis-driven, and evidence-led approach to crime reduction. Its focus is
towards catching offenders more quickly and getting upstream of the problems presented by crime by working to prevent it in
the first place. But how does crime science relate to crime mapping, and how does it fit in contemporary policing and crime
reduction? This presentation will describe the basis of crime science, how it complements the new paradigms in policing and
crime reduction, and how it is spiriting new forms of spatial analysis and crime mapping. The presentation will also describe
the profile it offers to crime analysis and the risks that will prevent its success.
John Jay College, CUNY
Geo-Spatial Analysis of Sex Offender Residency Restrictions in the State of New Jersey (Thursday: 1:30 - 3:00 p.m.)
Current trends in policymaking and legal/legislative issues are expanding on 1994 and 1996 federal legislation and are now
mandating that sex offenders live 1,000 feet away from any school. This paper focuses on the roughly 1,600 registered sex
offenders in the state of New Jersey and the distribution of their reported residential addresses against a mapped network
of schools and the 1,000-feet restriction buffer zones. The geo-spatial analysis includes high-density metropolitan and urban
area restricted living conflicts and the potential adverse effects the residency restriction policy would have on the New
Jersey sex offender population.
The Urban Institute
The Impact of Fear of Crime on Walking Outdoors: How Gang Territories, Violence, and Disorder (Thursday: 3:30 - 5:00 p.m.)
The current paper examines the environmental factors associated with fear of crime’s influence on physical activity, in this
case, walking outdoors. The research tests two hypotheses: (1) violent crime, drug activity, disorder, and the presence of
gangs decrease residents’ likelihood of walking outdoors due to fear, and (2) greater collective efficacy, neighborhood cohesion,
and other social capital constructs act as mediating factors to reduce the impact of violent crime, drug activity, disorder,
and gangs on fear that drives residents indoors.
Sarah M. Chilenski
Penn State, Prevention Research Center
Community Risks in Rural America: What Matters? (Friday: 3:30 - 5:00 p.m.)
The present study used GIS to aggregate agency-level reported crime data to the level of the school district. Analyses demonstrated
that school district crime rates had appropriate construct validity (e.g., related to economic risk, mobility, and population
density). Additional multilevel analyses indicated that the risk that high-crime areas present to adolescents operates on
a smaller level; agency-level (i.e., municipality) crime rates were more consistent predictors of youth problem behaviors.
Results indicate that district-wide prevention initiatives might be more effective if they recognize and address the differences
in levels of community risk in different sections of the district geographic area.
Heinz School, Carnegie Mellon University
Impact of Public Housing Demolition on Crime in Allegheny County (Friday: 10:00 - 11:30 a.m.)
Partly to ameliorate the effects of concentrated disadvantage, public housing initiatives during the 1990s relocated households
from older, high-density public housing communities to communities in the private housing market. The analysis documents similarities
and differences between origin and destination communities. Recent historical patterns in neighborhood violence levels provide
a basis for developing expected violence levels absent the housing relocation initiative. These will be compared with observed
outcomes to search for discontinuities in trends associated with the timing of major housing relocations. The results will
have implications for assessing the efficacy of housing relocation within current housing markets for violence prevention.
Kevin M. Curtin
University of Texas at Dallas
A Family of Models for Determining Optimal Police Deployments (Friday: 10:00 - 11:30 a.m.)
A family of optimal location models is presented that is designed to provide optimal patrol, deployment, and administrative
patterns for a range of police activities. A median model is presented as a tool for special team response; a dispersion model
is presented as an appropriate response for maximizing asset safety; a center problem demonstrates the worst case response
situation; flow covering models are presented to identify optimal network stop locations (such as inspection stations or roadblocks
for interdiction); and lastly, maximal covering models are presented as appropriate models for the equitable and efficient
location of administrative and patrol areas.
Allegheny County Department of Human Services
Spatial Analysis of Returning Offenders and their Use and Proximity to Services (Friday: 10:00 - 11:30 a.m.)
The Allegheny County Reentry Mapping Initiative integrated corrections (prison, parole, and jail) data with human service
client and facility data. This presentation will describe the results of this analysis with attention paid to: (1) What communities
are offenders returning to? (2) Are offenders booked in the jail returning to different communities than offenders incarcerated
in state prison? (3) Where are offenders in relationship to key human service supports? (4) Are there spatial differences
in those who access human services from those who do not? (5) Are offenders’ addresses changing from time of incarceration
to time of release?
Spatial Data Analytics Corporation
Transferable Spatial Signatures for Crime Forecasting (Friday: 1:30 - 3:00 p.m.)
The spatial signature extraction method has been extended to allow the effects of local environment to be filtered out such
that offender preferences can be transferred from one locality to another. This is useful for mobile criminals, criminal trained
in one area and active in another, or criminal networks that may operate in similar manners but are geographically distinct.
The presentation will include a statistical derivation of the model as well as implementation and example practical applications.
Virginia Commonwealth University
MOBESE: Crime Prevention through Technologically Integrated Design in Istanbul, Turkey (Friday: 1:30 - 3:00 p.m.)
MOBESE project has been selected as “The Best Mobile Application” of nationwide projects in 2001-2002, Turkey. This project,
then, has been improved and integrated by various technologies so as to meet the needs on security and crime measurements
in Istanbul, Turkey. This study represents MOBESE as an integrated design with MIS (Management Information System) & GIS (Geographical
Information Systems) including various spatial and surveillance technologies so as to enhance the communication amongst the
law enforcement units and to make information flow more reliable and faster. MOBESE, ultimately, aims to enhance crime prevention
processes through technologically integrated design in Istanbul, Turkey.
Middle East Technical University
Spatio-Temporal Analysis of Theft Incidents in Cankaya and Bahcelievler Police Districts, Ankara (Thursday: 1:30 - 3:00 p.m.)
Space-time clustering of theft incidents in Cankaya and Bahcelievler Police Districts of Ankara are analyzed based on exploratory
spatial data analysis methods and Knox index. Then detected space-time clusters are related to land use properties of the
region with the use of GIS techniques, and the time interval for the data is months. The monthly change in the pattern of
the theft incidents is explored by a series of two- and three-dimensional kernel estimates. It is detected that the theft
incident clusters move from the western part of the study region to the eastern part in the first half of the investigation
Safa F. Egilmez
Santa Monica Police Department
Exploratory Analysis of Homeless-Related Crimes and Social Service System in Santa Monica, California (Thursday: 3:30 - 5:00
GIS is used to analyze the spatial characteristics and patterns associated with homeless-related crimes and the existing and
planned location of homeless social programs. Use of GIS enables an exploratory data analysis that allows for understanding
the extent of the crimes committed by/on the homeless and gives decisionmakers a chance to situate future social programs
where they are needed. This research provides a starting point to understanding the importance of spatial planning of social
services in an urban area with a high concentration of homeless persons.
West Virginia University
“I Wasn’t There”: Space-Time Prisms for Criminal Investigation (Thursday: 1:30 - 3:00 p.m.)
Time-geography technique models of potential accessibility are applied to the movements of possible offenders. In combination
with route-finding and service area algorithms, a coded transportation network calculates the area potentially accessible
to an individual, which is displayed in two and three dimensions. Predicted behavior or past behavior is modeled based on
suspect’s known positions at given times and places. Predicted time-space paths are intersected with locations of known incidents
to provide information relating to the inclusion or exclusion of suspects from consideration. Examples are presented from
Morgantown, West Virginia.
University of Massachusetts Lowell
A Spatial and Contextual Effects Analysis of the Causes and Correlates of Law Enforcement Officers Killed and Assaulted in
the Line of Duty (Friday: 10:00 - 11:30 a.m.)
Using an opportunity theory and routine activities theoretical approach, this research examines the variations in the spatial
dynamics of assaults and killings of law enforcement (LEOKA) incidents across block groups within cities and the role of the
ecological and situational factors in accounting for variations in LEOKA incidents. Incorporating recent changes to the structure
of NIBRS introduced by the FBI in 2003, the research uses address-specific NIBRS data for the years 2003–2005 for nine cities
in Massachusetts supplemented with U.S. Bureau of the Census data for the year 2000.
Heinz School of Public Policy and Management, Carnegie Mellon University
Automatic Detection of Crime Frequency Pattern Changes (Thursday: 3:30 - 5:00 p.m.)
As the volume of crime data collected multiplies, it is increasingly important to automate the identification of interesting
phenomena in the data. We compare several methods’ abilities to automatically detect changes in average level of crime or
one-time outliers in the data by crime type at the car beat level of analysis. The judgments of three crime analysts on the
locations of these pattern changes in crime frequency–time series are used to train and evaluate the methods. We find that
monitoring percent change in crime frequency, a common practice, substantially underperforms methods that incorporate a univariate
Transport for London
Developing Aoristic Network Analysis on London’s Transport System (Thursday: 3:30 - 5:00 p.m.)
The complexity of London’s transport network and the way passengers use it both present complex challenges in analyzing non-static
crime events to improve policing and ensure passenger safety. Aoristic network analysis of London’s transport network is being
developed as a new tool to identify “hot sections” of the network that are not spatially definable using traditional analytical
techniques. A pilot study of theft on the Underground is being undertaken prior to extending to other transport modes (bus,
overground trains, and trams). This presentation will explain the challenge, approach, successes, and limitations to aoristic
Edmonton Police Service
Prostitution, Displacement, and Human Geography: The Affects of Police Actions (Thursday: 3:30 - 5:00 p.m.)
In response to public pressure concerning prostitution, the Edmonton Police Service launched “Operation 118 Ave.” The analysis
to be presented examined the impact this operation had on the location of prostitution activity. It is theorized that conventional
police initiatives will do very little to influence the spatial distribution of prostitution. Access to clientele and the
sex trade works perception of a safe environment plays a significant role in the location of a “stroll.” Despite police activities,
the spatial distribution of prostitution is not affected. Zones of tolerance could provide one solution to the prostitution
problem in Edmonton.
Institute for Law and Justice
Simulating Street Robbery to Inform Policy and Prevention (Friday: 1:30 - 3:00 p.m.)
Research examining the crime event in its context has been handicapped by a lack of micro-level data that can capture the
dynamic interactions of individuals and the context in which they occur. This research creates a simulation model of street
robberies that is based on extant theory and empirical research. The model is used to explore the impact of patrol allocation
and the presence of place managers on the spatio-temporal distribution of street robberies. Research conducted using simulation
offers a cost-effective supplement to field research because it allows the identification of the most promising strategies
for additional funding.
Department of Geography, Indiana University
Geographic Exclusion: Spatial Analysis for Evaluating the Implications of Megan’s Law (Thursday: 1:30 - 3:00 p.m.)
In Ohio, like many other states, Megan’s Law prohibits a sex offender from establishing or occupying a residence that is within
1,000 feet of any school-related properties. Critics believe that the geographic constraints placed on registered sex offenders
by Megan’s Law have the potential to eliminate most, if not all, feasible residential housing options. The purpose of this
paper is to outline a methodological framework for evaluating the impact of Megan’s Law and its potential implications for
geographic exclusion. A case study of Hamilton County, Ohio, is conducted, and empirical results are presented.
Karen L. Hayslett-McCall
University of Texas at Dallas
Simulating the Journey to Residential Burglary (Friday: 10:00 - 11:30 a.m.)
Our research presents an innovative, predictive approach to the study of the journey to residential burglary. We propose a
simulation model in conjunction with both routine activity and social disorganization theories. We will model the offender
and the residential location targeted for burglary as agents in a cellular automaton model. The model will use the reports
database from the Dallas Police Department to identify attributes of both the offender and residential location agents. This
will be supplemented by neighborhood information, location, and distance information derived through GIS techniques to model
the journey to crimes of residential burglary.
Cedrick G. Heraux
University of Michigan - NACJD
The Neighborhood Context of Police Use of Force Behavior (Friday: 10:00 - 11:30 a.m.)
The proposed research seeks to examine the effects of neighborhood context on the use of force using observational and survey
data collected in six different jurisdictions. By using neighborhood-level variables and available spatial analyses, this
research is better able to examine the effects of particular places on police behavior in addition to the variables traditionally
explored. This focus on the importance of context provides a new avenue for research on this topic and is viewed as a basis
for future studies of police behavior.
NACJD’s Role in Mapping and the Data Resources Program (Friday: 1:30 - 3:00 p.m.)
This presentation will describe the GIS and related data resources at the National Archive of Criminal Justice Data (NACJD).
The presentation will also focus on NACJD’s development of a geospatial data depository of crime data, discuss data security
and confidentiality issues faced in distributing these data, and describe GIS meta-data standards and formats established
to deliver information to researchers and policy makes.
Use of GIS for Identifying Problematic Areas or Officers (Saturday: 8:00 - 9:30 a.m.)
This research uses GIS to examine and display incident and arrest data to identify areas of a city where police are disproportionately
“disrespected.” Specifically, the crimes of fleeing and eluding police, resisting a lawful arrest, and obstructing an officer
will be examined separately then jointly to demonstrate how GIS can assist police managers in identifying problematic areas
or officers. This information can in turn be used to revise policies or practices to lessen underlying problems citizens might
have with the police that lead them to disobey police authority.
John Jay College of Criminal Justice
Perfecting the Denominator: Mapping Population Distribution in the Urban Environment (Thursday: 3:30 - 5:00 p.m.)
This paper discusses the importance of determining an accurate depiction of total population and specific sub-population distribution
for urban areas in order to develop an improved "denominator," allowing for more correct rates in GIS analyses involving spatial
modeling and crime analysis. Rather than using data aggregated by arbitrary administrative boundaries such as census tracts,
we use dasymetric mapping, an areal interpolation method using ancillary information to delineate areas of homogeneous values.
We review previous dasymetric mapping techniques (which often use remotely-sensed land cover data) and contrast them with
our technique, a Cadastral-based Expert Dasymetric System (CEDS), which is particularly suitable for urban areas. The CEDS
method uses specific cadastral data, land use filters, modeling by expert system routines, and validation against various
census enumeration units and other data.
Minneapolis Police Department
Spatial Aspects of Probationer Contacts with Police in Minneapolis (Friday: 1:30 - 3:00 p.m.)
The MPD has built an Automated Email Alert System that notifies individual probation officers daily when parolees on their
caseload have contact with the MPD. This system ensures timely notice to probation officers throughout Minnesota of the involvement
of their probationers in criminal events in Minneapolis, regardless of the probationer’s role in the event. A year of past
contacts was mapped to explore and identify concentrated areas of probationer contact. Spider map analysis was used to explore
the geographic distance between parolee residence and concentrated contact areas. The implications for police patrol and probation
services are discussed.
Office of Juvenile Justice and Delinquency Prevention (OJJDP)
OJJDP SMART System (Friday: 10:00 - 11:30 a.m.)
The Office of Juvenile Justice and Delinquency Prevention (OJJDP) has been developing the capacity to visualize available
resources that address multiple risk and protective factors which may be affiliated with juvenile delinquency and its prevention.
Using geospatial data, OJJDP has been successful in mapping many of its resources, including grants, promising programs, as
well as other Federal agency resource data with regard to various socioeconomic, crime, and other risk factor data, which
may be indicative of delinquency. Visualizing and reporting changes in crime patterns and socioeconomic factors over time
in areas with government-funded programs may reveal a program’s effectiveness/ineffectiveness.
The “Near Repeat” Phenomena of Dwelling Burglary Crime in Bournemouth, England (Friday: 10:00 - 11:30 a.m.)
Following published research on the communicability of dwelling burglary crime spatial analysis identified the “near repeat”
phenomena in Bournemouth and was used to support an intervention strategy aimed at reducing crime. Eventually, the scheme
was evaluated using time-series analysis and CrimeStat III to visualize the spatial change in dwelling burglary crime in the
intervention localities. Geo-demographics were used to find further residential areas into which the intervention could be
expanded. Further work now examines how this could be related to other crime types such as auto crime.
Portland Police Bureau
Portland Police Bureau Freeway Mapping Project (Thursday: 3:30 - 5:00 p.m.)
The Portland Police Bureau’s Freeway Mapping Project involved agency-wide collaboration in accurately mapping offenses on
Portland’s freeways. Prior to the project, these incidents were mapped to one of six points, one for each of Portland’s six
major freeways. This method rendered these data useless for crime analysis and mapping purposes and for policy decisionmaking,
such as in which neighborhood, precinct, and/or district to place these incidents. This presentation will describe the agency-wide
effort undertaken to implement this project and mapping techniques used to deliver a data set to be incorporated into the
Portland Police Data System, the bureau’s mainframe database.
New York State Division of Criminal Justice Services
Does Neighborhood Matter in Police Processing of Calls for Service? (Saturday: 8:00 - 9:30 a.m.)
Calls for service (CFS) data have frequently been analyzed to study the quality of life and crime. Understanding measurement
errors involved in CFS data is therefore of great significance. Based on CFS and crime incident data from a mid-size US city,
this research attempts to explore a potential bias in analyzing CFS data. Specifically, the research asks whether, and the
extent to which, neighborhood characteristics account for call inaccuracy. Results from spatial regression models and policy
implications will be discussed. Preliminary findings suggest that calls from a neighborhood characterized by low-income levels
and concentration of elderly residents tend to exaggerate the nature of events.
John Glenn School of Public Affairs, The Ohio State University
Dynamic Spatial Models of Fraudulent Behavior: GIS and Agent-Based Models (ABM) (Friday: 1:30 - 3:00 p.m.)
Our purpose is to present frameworks that can link GIS and ABM as a means for advanced spatial analysis. We apply the framework
to fraudulent behavior. GIS is used for visualizing spatial patterns that emerge from empirical data, whereas ABM is developed
to replicate such patterns by modeling the underlying processes. This combination of tools allows the testing and visual display
of the effect of policy options. We will discuss how this approach can enhance current uses of GIS, especially for decisionmaking
in public agencies.
Minnesota Department of Corrections
Cartographic and Statistical Analyses of the Effects of Double-Bunking Close-Custody Offenders (Friday: 10:00 - 11:30 a.m.)
In 2004, the Minnesota Department of Corrections experienced the largest 1-year increase in offenders when compared to all
other states, resulting in double-bunking a close-custody correctional facility. Noticeable increases in the number of rule
violations occurred. Through cartographic and statistical analyses, several variables were examined, including outdoor air
temperature, spatial density, age, race, gang membership, governing sentence, and years until release. Using GIS technology,
these variables were mapped throughout the correctional facility, assisting in the analysis and determination of variable
correlations. Results of this research project provide an explanation for rising rule violations and new offender management
City of Wichita
Using Hot Spots: Development of a Web-Based Interactive Tool for Command Staff Meetings at the City of Wichita (Thursday:
3:30 - 5:00 p.m.)
In the summer of 2005, the Wichita Police Department constructed a multifaceted briefing room for use with press briefing,
presentations, and command staff meetings. For command staff, the Department wanted a way to interactively analyze increases
and decreases in crime activity in a map environment. The City of Wichita GIS was used for development of the GeoCrime Command
Staff Application, an interactive, web-based hot-spot application.
Edmonton Police Service
Beware the Iron Grip of CAD: A Patrol Sergeant’s Perspective on the Utility of Crime Mapping in Frontline Policing (Friday:
1:30 - 3:00 p.m.)
Although academic discussions about crime mapping are becoming increasingly specialized, its practical application within
frontline policing remains unclear. This presentation seeks to examine how the patrol function’s symbiotic relationship with
computer-aided design (CAD) acts as an impediment to the full application of crime mapping in frontline policing. A variety
of other implementation issues detrimental to the long-term success of crime mapping initiatives in the patrol environment
will be examined, including the unreliable nature of CAD data, a pervasive case file mentality that defines a patrol officer’s
cognitive map, and how the fractured nature of patrol time itself acts as a disincentive, breeding indifference toward analytical
Nancy La Vigne
Justice Policy Center, The Urban Institute
Residential Mobility Among Returning Prisoners: Implications for Policy and Practice (Friday: 1:30 - 3:00 p.m.)
State and local prisoner reentry partnerships are increasingly relying upon maps depicting concentrations of released prisoners
to drive policy and practice on the ground. However, these maps typically rely on the addresses inmates provide to the Department
of Corrections upon release from prison, raising questions about whether the residential mobility of released prisoners might
render such maps inaccurate. This presentation will draw from interviews with 379 prisoners released from Texas correctional
institutions, for which number of residences and exact address locations were collected twice over the course of a year after
release. The presentation will discuss how often released prisoners move, why they do so, and how mobility affects research
findings and corrections practices.
Brian A. Lawton
Sam Houston State University
Assessing the Impact of Hurricane Katrina on Space-Time Clusters of Crime Patterns in Houston (Thursday: 1:30 - 3:00 p.m.)
Following the devastation caused by Hurricane Katrina, more than 250,000 New Orleans residents were relocated to the Houston-Galveston
area. Although there was an initial outpouring of sympathy and support for these evacuees, it has faded over time and most
recently resulted in evacuees being blamed for a citywide increase in the violent crime rate. These analyses examine patterns
of crime prior to and following the Katrina disaster. Particular focus is placed on examining geographic patterns of both
violent and drug crime and identifying trends that might be attributed to the displacement of the population caused by the
Louisiana State University
The Impacts of Hurricanes Katrina and Rita on Public Safety in Louisiana: A Spatial and Temporal Analysis of Reported Crimes
(Thursday: 1:30 - 3:00 p.m.)
This research analyzes the impacts of Hurricanes Katrina and Rita on crime rates in Louisiana and selected urban areas outside
of Louisiana. Reported crime data have been collected before, during, and after both hurricanes hit the U.S. Gulf Coas, to
assess spatial and temporal changes in the crime rates. Methods include exploratory spatial and temporal data analysis, including
co-maps; detection and mapping of crime hot spots and their changes over time; and spatial regression models to analyze the
relationships (and their changes) between crime and other demographic and socioeconomic data believed to be associated with
Assessment and Evaluation of Individually Calibrated Journey-to-Crime Geographic Profiling Models (Friday: 3:30 - 5:00 p.m.)
Criminal geographic profiling is a decision support tool used by law enforcement to make estimates about the likely location
of a serial offender’s “haven,” which, in most cases, is the offender’s residence. Modern geographic profiling models include
Dragnet©, RigelTM Criminal Geographic Targeting, Predator®, and the CrimeStatIII journey-to-crime and crime travel demand
Ned Levine & Associates
Bayesian Journey-to-Crime Modeling (Saturday: 8:00 - 9:30 a.m.)
A Bayesian approach to journey-to-crime modeling is examined. An estimate of the likely origin of crimes committed by a serial
offender can be created using the empirical Bayes (EB) method by adding estimates based on a trip distribution matrix and
a travel function calibrated from a known sample (journey-to-crime method). An empirical test of 50 serial offenders from
Baltimore County demonstrated that the combined estimate is more accurate than either the journey-to-crime method or trip
distribution method by themselves. The study has implications for combining information from multiple sources in a systematic
Texas State University – San Marcos
Illegal Drug Possessions in Urban Built-Up Areas (Friday: 1:30 - 3:00 p.m.)
This presentation will report on a study of the spatial patterns of illegal drug possessions in the city of Austin, Texas.
GIS is used to not only map the arrest locations of illegal drug possessions but also to spatially relate the offense locations
to those of property crimes and to land use and socioeconomic characteristics. Furthermore, exploratory spatial analyses are
conducted to evaluate the possible “attractiveness” that different types of urban facilities and socioeconomic activities
might have for illegal drug possessions. The goal of this research is to promote further explanatory studies of the location
factors of illegal drug possession.
Department of Geography, Indiana University
Spatio-Temporal Methods for Identifying Crime Hot Spots: Approaches, Problems, and Prospects
Innovations in crime mapping and hot-spot detection have focused on the integration of spatial statistical techniques and
GIS to improve the identification of elevated crime densities in urban spaces. Although such innovations assist in the development
of proactive policing strategies, these hot-spotting techniques fail to address the nature of hot-spots in both space and
time. This paper seeks to bridge this methodological gap by exploring the application of spatio-temporal methods for identifying
and analyzing crime hot spots. Empirical results suggest spatio-temporal analysis is a promising step in the search for more
accurate methods to identify crime hot spots.
Wayne State University
Mapping Out Prisoner Reentry in Wayne County, Michigan: A Gap Analysis Study (Friday: 1:30 - 3:00 p.m.)
In March 2006, the Hudson-Webber Foundation provided a grant to Wayne State University to conduct a gap analysis of prisoner
reentry in Wayne County. Results from the gap analysis provided information on (a) the geographic concentration of parolees
and probationers in Wayne County and the city of Detroit, (b) the distribution and overall capacities of key social and human
service providers, and (c) opportunities for enhanced coordination of community development initiatives and resources that
might positively impact prisoner reentry and overall community safety. The results of the study are currently being used to
guide the ongoing development of the Michigan Prisoner Reentry Initiative in Wayne County.
David C. May
Eastern Kentucky University
A Spatial Analysis of Predictors of the Relative Severity of Prison (Friday: 10:00 - 11:30 a.m.)
Previous research has revealed gender and racial differences in offender preferences for prison compared to alternative sanctions
and in the severity ranking of criminal sanctions among offenders. Nevertheless, both the opinion of the general public and
any spatial consideration of public opinion regarding perceptions of severity of prison have often been ignored. In this study,
the literature is extended by a spatial examination of demographic and structural predictors of Kentucky residents’ perceptions
regarding the amount of alternative sanctions that they would serve to avoid 1 year of imprisonment.
Measuring Crime Clusters Around Criminogenic Places: An Enhanced Buffer Approach (Saturday: 8:00 - 9:30 a.m.)
Research shows that crime clusters around certain types of land uses, including bars, schools, and transportation stops. The
accurate measurement of this phenomenon is central to crime theory and reduction efforts. We present an enhanced clustering
approach that is more sensitive to the proximity of crime events and land use locations than traditional buffers. It also
allows for a more realistic base rate when comparing numerous study areas and allowing the use of statistical methods. We
demonstrate both technique and statistical approach with a case study of drug arrest clusters around pawn shops in Philadelphia,
James W. Meeker
University of California Irvine
Using GIS to Study Legal Needs Issues: An Analysis of the LSC OIG Southern California Mapping Project (Saturday: 8:00 - 9:30
We evaluated an innovative effort by the Office of the Inspector General (OIG) of the Legal Services Corporation (LSC) to
explore the utility of using GIS to analyze the distribution of services for five legal service providers in Southern California.
This project is an extension of an earlier effort to employ this technology for the legal service providers in Georgia. The
application of GIS to analyze the delivery of legal services to the poor is a fairly new application. Many in the legal services
community are unaware of both the strengths and weaknesses of this analytical approach, and one of the major purposes of this
project is to increase awareness.
Exploratory Spatial Analysis of Juvenile Delinquency and Recidivism (Saturday: 8:00 - 9:30 a.m.)
This research investigates spatial patterns of juvenile delinquency and recidivism in Philadelphia and identifies relevant
spatial variation in the factors that might influence delinquency and recidivism. The application of global and local autocorrelation
statistics suggests that the spatially explicit causes of recidivism differ from those of delinquency and that different causal
factors might be operating in different regions. Results from a neural network-based sensitivity analysis show that the most
relevant attributes for predicting recidivism differ among the different regions of Philadelphia.
Exploratory Spatial Analysis of Juvenile Delinquency and Recidivism - cont. (Saturday: 8:00 - 9:30 a.m.)
The goal of the proposed project is to develop, apply, and evaluate improved techniques to investigate the simultaneous effects
of neighborhood and program forces in preventing juvenile recidivism. The analysis of neighborhood effects in this context
introduces a number of challenges, including issues of data quality, data integration, and the use of appropriate statistical
techniques. We address these challenges by using GIS and spatial data mining in an analysis of adjudicated juvenile delinquents
assigned to court-ordered programs by the Family Court of Philadelphia, Pennsylvania. The proposed study, which includes a
vast methodological departure from current practices, can greatly improve the chances of learning more about the dynamics
of juvenile recidivism and lead to more effective prevention policies and programs.
Jacqueline Warren Mills
Louisiana State University
The Clearinghouse Concept: Applications of Spatial Data for Law Enforcement and Public Safety (Thursday: 1:30 - 3:00 p.m.)
In a disaster event wherein mass evacuation has occurred, an essential question to the recovery process is “where have people
gone?” To reallocate resources to serve the redistributed population, this question must be answered to a suitable scale.
As a solution to integrating the myriad post-event geo-spatial data sets and answering this and other essential questions,
the LSU GIS Clearinghouse Cooperative (LGCC) was formed through collaboration between the Federal Emergency Management Agency
and LSU. This paper outlines the formation of LGCC, its utility in dealing with post-disaster geo-spatial data, and specifically
its application for post-event law enforcement and public safety concerns.
Office of Juvenile Justice and Delinquency Prevention (OJJDP)
OJJDP SMART System (Friday: 10:00 - 11:30 a.m.)
OJJDP has been developing the capacity to visualize available resources that address multiple risk and protective factors
that could be affiliated with juvenile delinquency and its prevention. Using geo-spatial data, OJJDP has been successful in
mapping many of its resources, including grants, promising programs, and other Federal agency resource data, with regard to
various socioeconomic, crime, and other risk factor data, which could be indicative of delinquency. Visualizing and reporting
changes in crime patterns and socioeconomic factors over time in areas with government-funded programs might reveal a program’s
John Derek Morgan
Florida State University
Visualizing Cases of Internet Fraud with Collaborative Crime Mapping (Saturday: 8:00 - 9:30 a.m.)
Internet fraud almost always occurs across jurisdictional boundaries that are not easily apparent to victims or institutions
effected by those crimes. Knowledge at the local and jurisdiction levels is specific to local actors (i.e., police officers
and citizens). This research proposes that a collaborative GIS using crime mapping driven by location-specific intelligence
can be effectively used to solve Internet crime cases. Existing efforts toward collaborative crime mapping will be considered.
Different implementation issues to such a proposed system will also be considered.
Thomas R. Mueller
California University of Pennsylvania (Cal U)
The Cal U Crime Mapping Center (CMC): Issues with Rural Data and Third-Party Crime Mapping (Friday: 1:30 - 3:00 p.m.)
Crime mapping has been used by police departments throughout the country to understand crime patterns and developing strategies
to combat crime. However some rural and small town police departments do not have the resources to implement such a plan.
The Cal U CMC was created to offer these services to small municipalities in the region. At the CMC, students input and analyze
data and then create weekly and monthly crime mapping reports. This presentation will examine the processes, successes, problems,
and movement to regionalism at the CMC.
John Jay College of Criminal Justice, City University of New York
Crime Specific Analysis of Homicide Hot Spots (Thursday: 1:30 - 3:00 p.m.)
Using 2005 New York City homicide data (n = 519), this paper, which is informed by a rational choice perspective, demonstrates
the importance of breaking down homicides into different types before undertaking hot-spot mapping. It also underscores the
need for detailed, systematic field observations to complement mapping if these hot spots are to be understood. Further, it
argues that the combination of mapping and field observations assists in identifying immediate and practical community crime
prevention measures. Finally, it discusses the importance of class projects such as this in helping crime mapping students
apply crime theories to prevention practice.
University of Florida
The Spatial Distribution of Arrests for Criminal Domestic Violence in Lexington County, South Carolina (Thursday: 1:30 - 3:00
An NIJ-sponsored outcome evaluation of the Lexington County Criminal Domestic Violence Court indicated that arrests for domestic
violence in Lexington County significantly increased after the inception of the specialized court in November 1999. The purpose
of the current research is to examine whether spatial patterns of domestic violence changed after the court was developed
and to examine community-level characteristics within Lexington County in areas that experienced a change in the spatial distribution
of domestic violence. A variety of spatial analysis techniques, including spatial data visualization, measures of local spatial
autocorrelation, and spatial regression, were used to explore these objectives.
New Mathematical Techniques for Geographic Profiling (Friday: 10:00 - 11:30 a.m.)
The main question in geographic profiling is: Given a series of linked crimes committed by the same offender, can we make
predictions about the anchor point of the offender? The anchor point can be a place of residence, a place of work, or some
other commonly visited location. Existing mathematical methods to calculate potential anchor points fall into two general
classes—centrographic measures and probability distance strategies. These are the mathematical methods used in existing geographic
profiling software. In our research, we developed new mathematical techniques to estimate the search area for the anchor point.
These new methods allow for the explicit incorporation of a range of geographical features, like jurisdictional boundaries.
Further, the underlying assumptions for these mathematical models are made explicit in terms of assumptions on offender behavior.
Texas A&M University
Measure of Risk Across Communities Due to Sex Crimes (Friday: 3:30 - 5:00 p.m.)
This paper explores how GIS can be used in solving issues related to sex offenders. Spatial analysis of sex crime and tracking
of registered sex offenders (RSOs) is shifting from simple geographical location analysis to the creation of risk management
and measurement models. The study presents the application of the critical risk zones model across communities and discusses
the applications for law enforcement practices and policymaking regarding sex crime. This research established a standardized
measurement of risk regarding RSOs that will permit law enforcement agencies to assess reduction of crime and existing risk.
Joseph E. Pascarella
New York Police Department
The Relationship between Distance Traveled By Offenders and Level of Violence in Shooting Incidents in New York City (Saturday:
8:00 - 9:30 a.m.)
Shooting Incidents (N=1,666) that occurred in New York City during 2004 and 2005 were analyzed to determine if there was an
association between distance traveled and level of violence to test core concepts of the Routine Activity Approach (Cohen
& Felson,1979) using a Journey to Crime (JTC) methodology. A non-linear regression analysis was constructed to determine the
log likelihood of the incident become fatal the further the offender traveled. There was an increase in level of violence
the further the offender traveled suggesting that distance traveled is a motivating factor in raising the level of violence
in a shooting incident.
Derek J. Paulsen
Eastern Kentucky University
Crime Series Analysis Software: A New Tool for Crime Analysts (Friday: 10:00 - 11:30 a.m.)
Over the last few years, there has been increasing interest in the use of expert systems in crime series investigations, with
particular focus on geographic profiling systems. However, existing geographic profiling systems have come under scrutiny
because of both their limitations and research into their accuracy. Specifically, traditional geographic profiling systems
are incapable of profiling commuter offenders, a significant portion of all offenders, and traditional systems’ overall accuracy
has been found to be no more accurate than simple centrographic methods. Because of this, research is beginning to look to
alternative ways to analyze crime series that go beyond traditional geographic profiling systems. In particular, this research
reports on a new method of analyzing crime series that assists in determining commuters from marauders and then provides the
optimal tool for analyzing the crime series. In addition to research results, discussion will include training and future
issues in crime series analysis.
Using Spatial Demographics to Improve Geographic Profiling (Friday: 1:30 - 3:00 p.m.)
One of the most talked about new technologies in the field of crime analysis is geographic profiling, an investigative methodology
that uses the locations of a connected series of crimes to determine the most probable offender anchor point. Although geographic
profiling software has been around since the early 1990s, the science of geographic profiling has languished considerably.
Much of the research on geographic profiling over the last 15 years has focused on developing new software platforms and providing
case study analysis rather than on critically evaluating issues of effectiveness and accuracy. This lack of research has created
a void in which many questions about geographic profiling and serial offending still exist. One such question is to what degree
can the analysis of opportunity structure help focus geographic profiling results? In particular, do offenders commit crimes
in neighborhoods that have spatial characteristics similar to their own residential neighborhoods, and if so, can this knowledge
improve profiling results? The focus of this research is on developing simple spatial diagnostics that can be used to improve
overall profile accuracy by helping to narrow sometimes large top-profile areas. In addition to research results, this presentation
will include a discussion of the implications of these findings for practitioners of geographic profiling.
Developing Standards and Curriculum for GIS in Law Enforcement (Friday: 1:30 - 3:00 p.m.)
This study analyzes and discusses data collected from police agencies currently employing GIS and crime mapping and from educators
who currently offer crime mapping courses in their curriculum across the country. Currently, there are no standards that guide
curriculum development in this area. In this study, law enforcement and educators collaborate to aid in crime mapping curriculum
development. Findings regarding what works and what doesn’t work in teaching crime mapping to police officers; agency-identified
needs provide a useful framework for developing GIS curriculum standards.
Ministerio de Seguridad de la Provincia de Buenos Aires
National Crime Mapping Program: Argentina’s First Experience (Friday: 1:30 - 3:00 p.m.)
This presentation deals with the start-up of a National Crime Mapping Program, including first steps taken in the actual implementation,
lessons learned, and recommendations for this kind of integration program. It also deals with data issues, cartographic problems,
and the unique environment of a Latin American country.
Port Authority of New York and New Jersey
Methodology for Aggregating Robbery Data (Saturday: 8:00 - 9:30 a.m.)
There are two common methods by which robbery incidents are aggregated to units of analysis, such as census block groups,
calculating rates, or using counts. Calculating robbery rates often poses a problem where a correctly measured denominator
requires having the population figure that represents pedestrian traffic in an area, and these data do not exist. A proposed
solution is to create a density surface map of the point data sets, overlay the block group polygons, remove the areas in
the study areas where the variables could not occur, and capture the mean value of the results of the robbery density surface.
Jerry H. Ratcliffe
Predicting Shootings from Near-Repeat Patterns: A Philadelphia Example (Saturday: 8:00 - 9:30 a.m.)
Property crime research has identified the near repeat phenomenon, where the risk of repeat victimization is transmitted from
one location to nearby places for a limited amount of time. This presentation explores the potential for near repeat patterns
to exist in a violent crime type: shootings. Using tools currently under development to quantify the spatio-temporal patterns
of near repeats, a study in Philadelphia finds elevated patterns of shootings within a short time and distance of previous
incidents. This space/time pattern can be quantified and mapped, and provides strong predictive intelligence on which to form
a shooting prevention policy.
Problems Associated With Bounding Spatial Data (Thursday: 3:30 - 5:00 p.m.)
Units of analysis are commonly differentiated between spaces and places. Places are generally depicted as points in space
while spaces have an aerial extent. There are a variety of problems and concerns that occur when point patterns of place data
are aggregated into space data. These problems and concerns are classified into problems associated with boundaries created
for other than research purposes, edge problems associated with spatially truncated data, modifiable aerial unit problem associated
with dispersion metrics changing as aerial boundaries change, and the dispersion of data and the appropriate statistical model
to use as aerial units change in size. Solutions to these problems are addressed.
Facility Cop: Data Management for Correctional Institutions (Friday: 10:00 - 11:30 a.m.)
The location in time and space of serious incidents in our nation's correctional facilities is a serious concern. Many correctional
facilities record this data in paper files that are cumbersome and difficult to retrieve. These paper files also are difficult
to aggregate by time, space, or by individual to determine when, where, and who are the particular problems in a correctional
institution. With funding from the National Institution of Justice, the authors developed a computer program to track incidents
within correctional facilities. This presentation highlights the capabilities of this program that is available free of charge
from the authors.
U.S. Department of Housing and Urban Development (HUD)
USPS Vacancy Data: A New Data Set for Measuring Abandonment and Blight (Thursday: 3:30 - 5:00 p.m.)
There is research that suggests abandoned or blighted properties might serve as a magnet for crime. Although most policymakers,
scholars, and neighborhood residents agree that abandoned and blighted properties are problematic, there is little consensus
on how to accurately define and count these properties. This presentation suggests a novel definition for abandoned or blighted
property and introduces new data set that uses U.S. Postal Service administrative data on vacant addresses. HUD’s Office of
Policy Development and Research is making this data available to researchers via Internet download.
State University of New York at Buffalo
Neighborhood Mismatch: Proximal Access to Mental Health Services for Youth (Saturday: 8:00 - 9:30 a.m.)
This paper is an examination of proximal access to mental health services for youth, emphasizing a neighborhood research approach
using census data and spatial analysis techniques. The purpose is to elucidate physical barriers facing youth in accessing
needed mental health services. Findings of this research are intended to inform the importance of need-based resource placement.
Lastly, the use of GIS and related spatial analysis techniques are currently underused in the field of social work, it is
hoped that the usefulness of this approach and the interdisciplinary activities of both practitioners and researchers will
realize success for underserved populations.
Appalachian State University
A Geographic Approach to Racial Profiling: Does the Location Explain Racial Disparity in Traffic Stops? (Friday: 3:30 - 5:00
In this study, the author analyzed the effects of spatial characteristics in patrolling areas upon racial disparity in the
frequency of traffic stops and subsequent police treatments. Through LISA cluster map analysis and spatial regression analysis,
it was found that the likelihood of being stopped and of being subjected to unfavorable police treatment (e.g., arrest, search,
and felony charge) was greater in beats with more blacks or Hispanics and/or more deployed police force.
Linda M. Rosato-Barone
City of Pittsburgh, Bureau of Police
MAPSTATS (Friday: 10:00 - 11:30 a.m.)
The lack of specific data regarding crime within a certain area hampers our officers’ ability to do proactive police work.
Although senior supervisors might have a handle on the crimes trends occurring within their area, unless it is filtered down
to the rank and file, it is of little assistance in reducing crime with proactive techniques. Allowing frontline supervisors
and officers access to current information in a timely manner gives them the ability to strategically map out a plan on a
daily basis to combat any crime trends that might be developing. MAPSTATS is just that tool, an intranet-based system for
analyzing aggregated incident reports geographically.
Birmingham Community Safety Partnership
Composite Index of Community Safety—Identifying Priority Neighborhoods Using Multiagency Data Sets to Create a Single Index
of Community Safety Mapped for Small Areas (Saturday: 8:00 - 9:30 a.m.)
Birmingham Community Safety Partnership used a composite index of multiagency community safety–related spatial information
to assess priority places. The paper demonstrates how the Partnership Information and Intelligence Team was able to combine
information from a variety of partner agencies to identify pinch points for multiagency delivery across the city. The output
was a comprehensive and innovative spatial strategic assessment of six priority themes that has shaped the ongoing strategy
of the key providers of community safety. The outcomes include a significant change in delivery policy with a move to neighborhood
management not only for community safety but also across all strategic partnership themes, including education, economy, and
Mapping Reassurance Using Survey Data (Thursday: 3:30 - 5:00 p.m.)
Birmingham Community Safety Partnership, through its Community Safety Mapping Online System, has created an innovative approach
to mapping survey data that allows users to look at reassurance and environmental hotspots in conjunction with incident-based
community safety data. The paper will share the cutting-edge methodology used and provide practical examples of the use of
the results in assisting strategic policy formulation and targeted response.
D. Kim Rossmo
Texas State University - San Marcos
The Error in Error Distance: Some Conceptual and Analytic Problems in Crime Mapping (Friday: 1:30 - 3:00 p.m.)
Crime maps are visual displays of quantitative information. Edward Tufte advises us of the importance of following fundamental
principles of analytic design and avoiding corrupt techniques in evidence presentation. Implications of his advice for the
field of crime mapping are discussed in this presentation, focusing on two examples: (1) the continued use of error distance,
despite its analytical unsuitability, in the measurement of geographic profiling algorithm performance; and (2) the introduction
of distortion and error in the condensation and packaging of crime analysis information for police practitioners, public safety
executives, the public, and the media.
Illegal Border Crossings by American Citizens (Friday: 3:30 - 5:00 p.m.)
This study examines the geographic and temporal patterns of American citizens apprehended while illegally crossing a U.S.
land border. Americans rank seventh in terms of citizenship for all illegal entrants and second in terms of citizenship for
criminal disposition entries (42% of American citizen apprehensions were criminal). An operationally valuable perspective
on border security can be gained by analyzing American citizenship cases. Individual characteristics for this group and the
geography of their entry points are examined and contrasted against similar findings for all illegal entrants.
Using GIS to Compare Regional Law Enforcement Information Sharing Systems (Saturday: 8:00 - 9:30 a.m.)
This presentation will cover the Comprehensive Regional Information Sharing Project application. This application uses ArcIMS
as an integral part of the interface for selecting and presenting data about six regional information sharing systems and
the participating police agencies. The presentation will cover how the application was developed, the current application,
and plans for further development. The presentation will conclude with a demonstration of the CRISP mapping tool.
Canadian Centre For Justice Statistics, Statistics Canada
Neighborhood Characteristics and Crime Distribution (Thursday: 1:30 - 3:00 p.m.)
This presentation will begin with an overview of the Canadian Centre for Justice Statistics and its overall objectives. The
presentation will show how crime mapping can be used to develop crime prevention strategies in the Canadian context.
Examining Crime using Statistics Canada Geocoded Data (Friday: 1:30 - 3:00 p.m.)
This presentation will highlight Statistics Canada’s involvement in spatial analysis of crime data. The Canadian Center for
Justice Statistics (CCJS) is a division of Statistics Canada. The Center operates a broad range of surveys across the major
sectors of the civil and criminal justice system including the Uniform Crime Reporting incident-based survey, courts, corrections
and victim based surveys. The CCJS is responsible for collecting and analyzing crime data using an integrated approach.
Trenton Police Department
Utilization of Crime Mapping in the City of Trenton (Friday: 3:30 - 5:00 p.m.)
The utilization of crime mapping affords law enforcement professionals the ability to analyze crime events with the benefit
of visualization not afforded in the printed word. Analysis of crime mapping data provides not only the ability to detect
geographic patterns of crime but also to develop strategies to abate the crime threat, evaluate the performance of operational
plans, and determine the need for additional strategy implementation. Crime mapping also provides law enforcement executives
with an accountability system that can determine if personnel within their agency are performing their duties as they have
Logistics and Quantitative Methods, CSIR Built Environment
Determining Criminal Activity Space Using Mobile Phone Technology (Thursday: 3:30 - 5:00 p.m.)
We propose two methods for determining a suspected criminal’s activity space using mobile telephones. The first uses call
data records from a suspect’s mobile telephone (the cell for each call made and received). The second involves actively tracking
the mobile telephone and obtaining its location at specified time intervals. For both, we use the cell locations and movements
between the cells to determine the most likely roads used by the suspect as they move around. The cell information and road
network help determine the activity space. Frequently used cells indicate possible anchor points.
The Research Applications for GIS in Identifying Demographic Changes Over Time (Friday: 3:30 - 5:00 p.m.)
Gentrification, urban renewal, and other types of neighborhood change are important areas of study for researchers interested
in urban crime patterns. GIS can greatly benefit the site selection phase of research projects that aim to capture the type
of micro-level phenomena that might be the most closely associated with crime (i.e., the density of social networks and availability
of local resources). This paper discusses several methods for visually presenting multiple years of census data to help identify
specific areas that have experienced dramatic demographic changes over time.
Redlands Police Department
GPS in a Crime Analysis Context: Practical Considerations, Research Needs (Friday: 3:30 - 5:00 p.m.)
Global Positioning System (GPS) data is hitting the crime and intelligence analysis profession from every angle—AVL, offender
monitoring, cell phones, etc. Whether these new data streams yield improved tactical and strategic analyses rests with how
GIS practitioners, crime and intelligence analysts, researchers, and our technology allies prepare for and adapt to this change.
An overview of GPS considerations, examples of how this data is being used, and suggested approaches to delivering today while
preparing for tomorrow’s avalanche will be discussed.
Justice & Security Strategies, Inc.
Building a Village One Map at a Time (Friday: 3:30 - 5:00 p.m.)
The ramifications for incarceration go beyond the individual felon. Incarceration has unseen consequences for children and
families. A multidisciplinary team is using GIS/spatial analysis to educate about the children left behind from incarceration.
Data from the criminal justice system are integrated with data from the social services sector to understand issues and redefine
social services asset mapping. The session will describe how GIS analysis is used to: (1) identify the need for services for
children of inmates, (2) acquire funds, (3) advocate with policymakers, and (4) assess the impacts of prevention efforts.
University of Edinburgh
A Learning Through Principled Discovery Framework of Monitoring Social Indicators: An Exploratory Spatial Data Analysis Approach
to Understanding the Context of Social Programs (Friday: 10:00 - 11:30 a.m.)
We describe an exploratory spatial data approach to understanding changes in key social indicators. Using an example of public
health surveillance data from Scotland, exploratory spatial data analysis (ESDA) methods are used to develop the notion of
a risk landscape as a measure of the context in which operates are discussed. Our approach applies a learning through principled
discovery approach for a centralized unit (such as a State) to monitor and learn from developments in localities. Links between
our approach and a realist framework of evaluation are discussed
Crime and Its Proximity to Drug Gang Sites: A Spatial Intelligence Challenge (Friday: 3:30 - 5:00 p.m.)
Traditionally, the void between criminal intelligence and crime analysis has kept drug and gang investigations separated from
the crime problems that they might cause. The case-specific nature of gang investigations has rarely merged well with the
high-volume focus of much crime analysis. In this paper, we use a new localized buffer technique to determine the intensity
of property and violent crime around drug corners associated with different gangs. The technique allows for better measurement
of crime in the vicinity of gang corners, and we find significant variation in the crime intensity for non-gang and gang locations
and disputed sites.
Jasper van der Kemp
Vrije Universiteit Amsterdam
Crime Patterns Reanalyzed: The Offender in Sight (Thursday: 3:30 - 5:00 p.m.)
Crime patterns are analyzed in various ways with different theoretical concepts and techniques. The focus of attention can
be directed at the city level or pointed toward the pattern of crime locations of an offender. From both neighborhood and
offender travel behavior studies, hypotheses can be drawn about factors of influence of these patterns, such as the target
backcloth, infrastructure, and general characteristics of neighborhoods. This paper will discuss how the city level is influenced
by the patterns of offenders, and a concept about the underlying factors will be discussed.
Asheley Van Ness
Massachusetts Executive Office of Public Safety
Moving Beyond Maps: Using Spatial Analysis Methods to Understand Violence and Reentry Issues (Friday: 3:30 - 5:00 p.m.)
This presentation will focus on moving beyond GIS modeling by applying various CrimeStat routines to violence and offender
reentry data. Spatial analysis techniques will be explored (as appropriate) to address violence and reentry policy questions,
such as neighborhood planning, service delivery planning, and employment issues. More specifically, we will use spatial descriptions
to look at where employment services for ex-offenders should be located, use distance analysis methods to examine if violence
is randomly distributed across a study area, and hot-spot analysis to identify clusters of returning offenders who do and
do not receive social services and areas in the community that may have higher concentration of violence.
Isaac T Van Patten
Radford University, Department of Criminal Justice
Spatial Configuration of Sexual Homicide in Los Angeles: A 25-Year History (Friday: 1:30 - 3:00 p.m.)
Twenty-five years of sexual homicides in Los Angeles County were examined from the context of routine activities theory. The
spatial configuration of 195 sexual homicide incidents occurring between 1980 and 2004 were evaluated using the techniques
of exploratory spatial data analysis. Victim-to-body dump and offender-to-body dump trips are then analyzed in terms of distance
traveled and directionality of travel. Finally, distance type mobility triangles are developed, and the incidents analyzed
using the Neighborhood, Predator, Intrusion, Offense-Mobility, Total Mobility typology developed by Tita & Grifiths (2005)
and extended by Groff & McEwen (2006). The results are discussed in the context of both technical issues in the analysis and
the implications for understanding the overlapping activity spaces of victims and offenders.
Irvine Police Department
The Use of Mapping for Tactical Decision-making in a Serial Burglary Case (Friday: 1:30 - 3:00 p.m.)
For several years, the city of Irvine experienced a high number of residential burglaries, and it was believed that a serial
burglar was at work. Crime analysts were tasked with determining the case linkage, detecting temporal and spatial patterns,
and providing tactical information for the development of an operation plan. This presentation details the role analysis and
mapping played in the creation of a tactical action plan that resulted in the identification and apprehension of the offender.
Northern Illinois University
Application of a New Clustering Method in Analyzing Homicide Patterns in Chicago (Friday: 10:00 - 11:30 a.m.)
The rate estimates for rare events like homicide in small populations are very susceptible to data errors and thus compromise
the validity of inferences. One approach is to construct geographic areas with sufficiently large base populations to permit
reliable estimates of homicide rates to be obtained. This presentation introduces a new clustering method based on the scale-space
theory that groups polygon objects of similar attributes together and thus creates larger analysis areas. The method is implemented
in a GIS program for wide distribution. A case study of homicide patterns in Chicago demonstrates effectiveness of the method.
San Diego County District Attorney's Office
Sex Offender Laws: Planning for an Election (Thursday: 1:30 - 3:00 p.m.)
Over the last year or so, many States and municipalities have implemented new, tougher sex offender laws. Modeled after Florida’s
Jessica’s Law, California put a similar law on the ballot for November. The San Diego District Attorney (DA) has been a big
proponent of this ballot measure, which has required a wide number of maps and geographic analyses. This presentation will
provide an overview of the various maps and analyses completed for the DA in anticipation of the election.
Using Spatial and Geo-Demographic Analysis to Understand Distraction Burglary in the East of England (Friday: 10:00 - 11:30
Distraction burglary is a particularly unpleasant crime where offenders prey on vulnerable people. The fact that offenders
are thought to travel extensively to commit their crimes means that confining analysis to one police force area will not fully
address the problem. This research aimed to investigate the patterns of distraction burglary, particularly exploring the interaction
between victims, offenders, and place. The research used GIS to explore the spatial patterns of offenses and to consider the
link between geo-demographic classifications, deprivation, and victims’ addresses. Strong links were found between geo-demographic
classifications and the chances of being a victim of distraction burglary.
Susan C. (Wernicke) Smith
Shawnee (KS) Police
Using GIS for Search Warrants (Thursday: 1:30 - 3:00 p.m.)
Local law enforcement has found many uses for GIS is tracking and monitoring crime incidents and movement and in forecasting
future incidents. However, an application of GIS in preparing for and executing search warrants can be crucial, as officer
safety, witness/bystander safety, and even suspect/arrestee safety is at its highest vulnerability during these events. This
presentation will provide examples of how GIS was integrated and proved to be crucial in the successful execution of three
search warrants conducted by Shawnee Police Department’s SWAT team. In all three examples, information discovered through
GIS was used to enhance the safety of all involved and protect evidence and property during the execution of a search warrant.
Using GIS Technology to Track, Forecast and Apprehend a Serial Robber in the Kansas City Metropolitan Area Across Multiple
Jurisdictional Boundaries (Friday: 3:30 - 5:00 p.m.)
In 2005, the Kansas City Metro area was plagued by an extremely active “fast food” robber. After striking across many jurisdictional
boundaries, Shawnee Police Department’s crime analyst was tasked with analyzing and forecasting possible hits for the entire
metro area. Using data gathered from the agencies, along with GIS information from a variety of police and non-police databases,
the analyst was able to identify a specific pattern related to the geography of the crimes- which included repeat victimization
- and forecast accurately the next hit in Independence, MO. IPD was surveilling the forecasted location when the robber struck.
He was taken into custody.
Idaho State Police
Challenges and Opportunities Rural States Face in Mapping NIBRS (Thursday: 1:30 - 3:00 p.m.)
Research was conducted to determine the best way to map National Incident-Based Reporting System (NIBRS) data in Idaho, which
has many counties with very sparse populations. Thematic maps showing crime by county yielded little if any usable information
because they were not able to show where Idaho crimes were actually occurring. Mapping NIBRS data by the jurisdictional boundaries
of city, county, and State agencies has shown positive rewards. County- and State-level data are mathematically estimated
and placed where they most likely occurred.
University of Florida
Repeat Single Family Burglary—Spatial and Temporal Patterns (Friday: 10:00 - 11:30 a.m.)
With the tools of spatial and temporal analysis, this research seeks to (1) explore the time course of repeat single-family
burglary, (2) compare spatial pattern of repeat single-family residential burglary and hot spots of all single-family residential
burglary (repeat and non-repeat), and (3) compare the findings with international studies. In addition, we explore some of
the improvements in methodologies that would enhance the study of repeat victimization. This study analyzes 3,100 residential
burglary cases reported by the Gainesville, Florida, Police Department over a 4-year period to explore repeat burglaries.
SMART + Kaleidoscope: National and Regional Analysis for Juvenile Justice (Friday: 10:00 - 11:30 a.m.)
Crime mapping has matured beyond a desire to simply map crime events. There is a broad realization that mapping crime is not
enough—rather, there is a need to analyze and visualize risk factors that lead to crime. Toward this end, the Office of Juvenile
Justice and Delinquency Prevention has developed the Socioeconomic Mapping and Resource Topography (SMART) system to provide
visitors with not only an event mapping capability, but also the ability to plot federally funded programs that address delinquency
and crime. Visitors can create maps and retrieve statistics using the socioeconomic, crime, and resource data provided at
various geographic levels, including the State, county, and local levels.
Date Created: January 28, 2008