The long-term goal of these and related activities is to develop a fully functional Crime Analysis System (CAS) with standardized data collection and reporting mechanisms, tools for spatial and temporal analysis, visualization of data and much more. Among the drawbacks of current crime analysis systems is their lack of tools for spatial analysis. For this reason, spatial analysts should research which current analysis techniques (or variations of such techniques) that have been already successfully applied to other areas (e.g., epidemiology, location-allocation analysis, etc.) can also be employed to the spatial analysis of crime data.
The following lists current problems that hamper the development and availability of crime analysis systems in the law enforcement community. I believe that these issues are not only restricted to crime analysis systems, but are also applicable to other areas of GIS and spatial analysis applications.
1. Money and time constraints
Police departments have limited resources to purchase computer hardware and GIS, spatial analysis, and mapping software. Training police officers with this new technology costs further money and time. Building a comprehensive crime database that can be implemented, shared, and updated among the different units in a police department costs additional time and money. Smaller police departments (less than 1000 employees), because of very limited financial resources, are especially handicapped to fully participate in this new development. On the other hand, larger police departments have already established their own crime analysis units including full time personnel to apply for research grant money (examples are New York City, Chicago, Washington D.C., Los Angeles, etc.)
2. Confidentiality, security, and accessibility of crime data
Crime data are originally collected at the individual level. Police reports usually record individual data (sex, age, race, address, etc.) on the victim, the offender (if apprehended), the crime location, and time of incidence. This involves important issues of data confidentiality, security, and accessibility.
3. Lack of training in GIS, spatial analysis, and computer mapping
Police officers and detectives lack training in these new technologies and might even be computer illiterate. For this reason any software product developed for law enforcement agencies needs to be easy to use, preferably with a point and click interface, steep learning curve, and appropriate default values. It is no surprise then that MapInfo and ArcView are the currently two most often used GISs among police departments in this country. Additionally, the CMRC has coordinated six different modules on crime mapping, the use of GIS, and spatial analysis that will be taught to interested law enforcement agencies nationwide.
4. Lack of spatial statistical software that is targeted specifically for crime analysis
The development of spatial crime analysis tools and their implementation into existing commercial GIS packages requires the combined effort of law enforcement agencies, academia, and software engineers. The perhaps first stand-alone crime analysis package is the so-called Spatial and Temporal Analysis of Crime (STAC) package, which was developed by the Illinois Criminal Justice Information Authority in the mid-80s. This software possesses various spatial and temporal analysis tools to detect patterns of crime in a community, using both geographic and time data (Bates 1987). The spatial analysis part is restricted to calculating the nearest neighbor index and the identification of crime hot spots. A second example stand-alone analysis package is POINTSTAT, a spatial statistics program for the analysis of point locations which can be used to analyze events (e.g., crime incidents, accident locations) or the spatial distribution of particular types of organizations (Levine et al. 1994). The most recent example is the so-called Point Pattern Analysis (PPA) software, developed by Chen and Getis (1998). A major drawback of these programs is that they are purely analysis tools and need to be loosely coupled with a GIS (e.g. Arc/Info, MapInfo) for displaying the analysis results (e.g., crime hot spots).
An example of an extension kit to an existing GIS is CrimeView by the Omega Group, a suite of integrated crime analysis tools designed for use in the object oriented ArcView GIS environment. Such crime analysis systems provide the full functionality of a GIS tailored to the specific needs of law enforcement agencies (e.g. standardized data collection and reporting mechanisms; tools for spatial and temporal analysis as well as visualization) and to the specific nature of crime data (i.e., point location in time).
Projects currently underway include the development of new GIS procedures for analyzing incident data (Levine and Wong 1997), for predictive modeling and for enhancing proactive policing (Hunt and Zubrow 1997). Current spatial point pattern analysis software that has been already applied to geographical epidemiology is summarized in Gatrell et al. (1996).
I am very interested to participate in the workshop on status and trends in spatial analysis. This workshop provides me with the opportunity to find out which new spatial analysis tools already applied to other areas of geographical inquiry might also be applicable to the spatial analysis of crime. It also gives me the opportunity to share with the workshop participants recent trends in spatial crime analysis, associated needs, and problems.
Together with one of my graduate students, I am currently developing computer programs that calculate the spatial association between two or more point data sets (e.g., crime locations, offenders and victims residences, etc.). Such techniques are applicable to a wide array of geographical problems. Another one of my graduate students currently develops models of spatial crime displacement due to urban renewal programs that have been carried out in the city of Baltimore, MD. Such models will help city planning bureaus and law enforcement agencies with pro-active decision making with the goal to decrease criminal activities.
Chen, DongMei, and A. Getis 1998. Spatial Point Pattern Analysis (PPA) Software. Department of Geography, San Diego State University.
Gatrell, A.C., Bailey, T.C., Diggle, P.J., and B.S. Rowlingson 1996. Spatial point pattern analysis and its application in geographical epidemiology. Trans Inst Br Geogr, NS 21:256-274.
Hunt, E.D., and E.B.W. Zubrow 1997. Building Crime Analysis Extensions for ArcView. Arc/Info User Conference. CA
Levine, N., Khoo, C., Okazaki, D., Kim, K., and L. Nitz 1994. Pointstat:
A statistical program for analyzing the spatial distribution of point locations
(version 1.0). Honolulu, HI: University of Hawaii at Manoa.