Julia B. Conley

Elizabeth Groff

Tim Lesser

Using GIS to Support Community Safety:
Strategic Approaches to Community Safety (SACS) Initiative
in Winston-Salem, NC

ABSTRACT: Winston-Salem, NC has been designated as one of five communities participating in the Department of Justice (DOJ) sponsored Strategic Approaches for Community Safety (SACS) Initiative. The Winston-Salem Initiative relies on DOJ, local agencies and researchers to provide and integrate diverse community data sources for the implementation of research-based strategies to reduce and prevent juvenile violence. GIS coverages and applications have been developed to support this integrated method . Some of our challenges related to the application and database development have included: identification, integration and standardization of data from multiple agencies, maintenance of confidentiality and security, and geocoding.


INTRODUCTION

The Department of Justice (DOJ), as part of a continuing effort to assist in developing strategies that reduce crime in communities, developed and is piloting the Strategic Approaches to Community Safety Initiative (SACSI). SACSI is designed to promote a multi-agency collaborative approach to data-driven problem solving. This paper focuses on one of four major components of SACSI, the development of the research and technology infrastructure in communities. Technology enhancements are viewed as necessary precursors to the project's strategic, data driven focus and have been given a high profile within the SACS Initiative. Since the project is still underway, the purpose of this paper is to provide early and ongoing documentation of challenges and lessons learned in one site, Winston-Salem, NC.

Strategic Vision of the Initiative

The Strategic Approaches to Community Safety Initiative (SACSI) was launched in early March, 1998, in an effort to test a specific framework for combating local crime problems (Concept paper, NIJ, 1998). Five cities are participating in the two-year pilot project: Indianapolis, IN; Memphis, TN; New Haven, CT; Portland, OR; and Winston-Salem, NC. These participants were selected from a pool of cities having the following characteristics: 1) a population of over 100,000; 2) a violent crime rate either above average or on the rise; 3) a police department with some degree of technological infrastructure; 4) an innovative and forward thinking Mayor, Police Chief and/or US Attorney; and 5) a history of, or capacity for, collaboration.

At its core SACSI is an effort to "increase the capacity of U.S. Attorneys, working in partnership with federal, state and local criminal justice agencies and a research entity, to collaborate on data collection and analysis, and design targeted strategies and interventions to prevent and reduce crime" (Concept paper, NIJ, 1998). The framework of SACSI has four major components: 1) formation of an inter-agency working group; 2) enhancement of a research and technology infrastructure; 3) use of a defined set of problem-solving process steps; and 4) transfer of the lessons learned from the project. Each of these components is described in more detail below. These components are being used as a guide as the five sites work through their two year pilot projects. However, the process in each site is very much a site specific and an organic one.

Previous research in Boston, Minneapolis and other jurisdictions has illustrated the importance of a multi-agency, collaborative approach to problem solving (Kennedy, 1997). Based on these examples, the first component of SACSI outlines the need for the formation of a strong inter-agency working group and details who should be a part of it. The "core group" has representatives from the following types of agencies: law enforcement, community, government and university. Members of this group work together to identify a specific crime-related problem, develop creative methods for addressing the problem, and evaluate the initiative's effectiveness and progress. A unique characteristic of SACSI is its emphasis on including non-law enforcement agencies in both the identification of problems and in the formulation of strategies to address those problems. Key members of the "core group" include the US Attorney, the project coordinator, and the research partner.

The second component, enhancement of the research and technology infrastructure, reflects the importance of data and the ability to synthesize data into knowledge. In order to make informed decisions about strategy, the team members need to have direct access to a wide variety of data. Once access is established, an easy-to-use interface between the data and the decision-makers needs to be created so the data can be utilized. Location is an important variable in a multi-agency effort to combat crime because it can serve as the common denominator among the various individual databases. In order to provide analytic capacity that includes spatial analysis, the integration of Geographic Information Systems (GIS) is required.

Another component of the Initiative requires the use of a problem solving approach to crime. Core team members are encouraged to utilize a seven step approach to problem solving: 1) identify the problem; 2) analyze the problem; 3) identify trends, patterns, opportunities for intervention; 4) design the strategy; 5) implement the intervention; 6) evaluate intervention; and 7) adjust the intervention. Each of these steps requires both access to accurate and timely data and the ability to analyze those data.

"Enhancing the Technology"

The information systems portion of the Strategic Approaches to Community Safety Initiative began in early April 1998 guided by the principle that we were going to "enhance the technology" at each of the five SACSI sites to provide the basis for a problem solving approach to solving crime problems. In keeping with the organic nature of SACSI, "enhancing the technology" continued to be refined until it was distilled into three major goals: 1) enable data access; 2) allow use by all team members; and 3) increase the spatial analytic capacity of the site. This portion of SACSI was directed by staff at the National Institute of Justice's (NIJ) Crime Mapping Research Center (CMRC).

To support the multi-agency, collaborative approach advocated in SACSI, participating agencies need to have access to their own data and other agency's data. Since agencies outside of the criminal justice system were included in the effort, access to both traditional criminal justice data and non-criminal justice related data would be necessary. The access also had to be from the agency rather than requiring a trip to some central location. Since SACSI is a collaborative effort, any system designed to support the Initiative must enable participants to access and use information from other agencies as well as their own data. However, all data sharing will be subject to the constraints of both legal and customary confidentiality regulations.

Beyond access, participants in the Initiative also need an easy to use method for analyzing the data. Raw data is not usually a problem in the field of criminal justice. However, gaining access to those data and then turning the volumes of data into knowledge constitute tremendous hurdles. Any "enhancement" to increase analytic capacity must include an easy-to-use front end that would allow SACSI participants the ability to utilize the data for problem solving. The information system developed must be easy-to-use so that all members of the team are able to conduct their own data analyses.

The third goal was to harness the tremendous power of geographic information systems (GIS) to enable the visualization and analysis of spatial data and make it available to SACSI participants. GIS combines both spatial and tabular analysis techniques with the additional advantage of being able to visualize the relationships among data sets. This capability allows visualization of one type of data (e.g., street robberies) and relationships between data sets (e.g. street robberies and drug markets).

In addition to the goals of "enhancing the technology" in the sites, there were also two guiding principles under which the implementation was conducted. The first principle concerned developing a system that reflected user needs in each site. Secondly, the new information system would build on existing efforts and reflect the software (both database and GIS) already in use at the site. Consequently, each site's version of the CSIS will be different because it will build on resources already in place and be driven by user defined preferences. The comprehensive and well-constructed needs assessment procedure utilized in each site ensures that these two principles are met.

Needs Assessment Phase

In order to meet the goals and principles outlined above we began by conducting a needs assessment in each site. The approach used by the assessment team consisted of a systematic methodology that has been utilized previously in numerous successful GIS-related user needs assessments (Calkins, 1996). This methodology was instituted to try and ensure the accurate characterization and documentation of user needs. Even though it is time-consuming, this step is necessary if we hope to implement systems to meet those needs. Additionally, DOJ/CMRC hired INDUS Corporation to conduct the needs assessments in two of the sites, assist with one site and provide support on all five sites. Dr. Calkins trained both CMRC and INDUS staff on his methodology. Use of the same methodology by both CMRC/INDUS teams and CMRC teams resulted in the necessary consistency across all five sites.

Conducting a formal needs assessment not only provides information, it also functions as an educational process and generates excitement about the project. As part of the interview process, participants are informed about GIS and how it can help support their project. Additionally, the interviewers often are able to pass along information about data sources that they have gleaned during previous interviews. Finally, by informing potential participants about SACSI, a higher level of enthusiasm for the project was generated.

SACSI DEVELOPMENT IN WINSTON-SALEM

The local issue of juvenile violence was chosen to be the focus of the Strategic Approaches to Community Safety Initiative here in Winston-Salem. The SACSI juvenile age criteria has also included 16 and 17 year-olds, even though they are considered adults under North Carolina law. After the needs assessment and parallel to the development of the GIS technical component, the core team of SACSI and front line workers from local law enforcement, local elementary, middle and high schools, social service, mental health and non-profit agencies serving our community's youth were meeting with local researchers to analyze data on juvenile and young adult violent crimes and the local nature of juvenile violence within three different age groups, 16-17 year-olds, 12-15 year-olds, and 11 years and under. This collaboration of local partners included the Winston-Salem Police Department (WSPD), Winston-Salem/Forsyth County School (WSFCS) System, Office of Juvenile Justice (OJJ), Forsyth County (FC) Sheriff's Department, CenterPoint Human Services, FC Department of Social Services (DSS), Department of Community Corrections, Forsyth District Attorney's Office and researchers from Wake Forest University (WFU), the lead research institution, and Winston-Salem State University (WSSU). Agency data from WSPD, FC Sheriff's Office, OJJ, DSS, CenterPoint, and WSFCS have been analyzed during this process.

SACSI Community Involvement and Local Data Analysis

One of the strengths of SACSI has been the data driven local knowledge and the input from those with first-hand experience working with juveniles. We have minimized making subjective assumptions with this approach. Prior to implementation of the GIS application component of the project, the researchers and front line workers have defined and focused on areas of the city with high juvenile crime rates. To provide the accurate data to assist in these efforts, numerous reporting analysis has been generated from WSPD data for juveniles involved in drug violations, violent crimes - homicide, rape, robbery, aggravated assault, weapons violation, kidnaping, and the early pathways - simple assault and communicating threats in younger cohorts 11 years and under.

The local juvenile crime involvement has been analyzed by cohort age ranges, by Fire Demand Zones (FDZs), FDZ groupings to define neighborhoods, police beats, sectors, by frequency of perpetrator involvement, etc. This has also entailed address matching the home addresses of all juveniles with involvement in these crime categories. The emphasis on reducing juvenile violence led to a focus on incident reviews for four initial geographic areas with the highest concentrations of crime. We also looked at adult criminal influence by age ranges in these neighborhoods. The incident review teams included beat officers, probation officers, FC School Resource Officers (SROs) and officers from WSPD Juvenile Repeat Offender Unit.

The computer software used at the WSPD to support the data analysis requests from SACSI includes Seagate Crystal reports 7.0, MS Visual FoxPro 6.0 and COBOL 74 , which have been used for data extraction and reporting. MS Visual FoxPro 6.0 was used for home address corrections and standardizations, and ArcView 3.1 and ARC INFO 7.2.1 were used to geocode and create thematic maps by sectors and of specific areas of the city with higher crime concentrations, by time of day, during school year, by juvenile home addresses, etc. This mapping and reporting has assisted the local efforts of SACSI to understand the nature of our juvenile crime and design strategies to combat the problems identified. It has also raised the awareness of the types of ongoing reporting and analysis capability that will be needed from the Community Safety Information System.

Jason Network

Another tool that is being developed in our community to support SACSI efforts to prevent juvenile violence is the Jason Network. This is a MS Visual FoxPro client-server based application, that provides for building a case management database around the most serious, chronic juvenile offenders and sharing the appropriate information among the agencies that are focused on monitoring and providing intervention strategies to redirect these serious juvenile offenders. This tool will help identify the chronic, as well as the emerging violent juvenile offenders in our community. It is currently being developed by Open Software Systems, Inc. using WSPD, OJJ and SACSI data sources. Access is limited to some of the same partners participating in the SACS Initiative, counselors in WSFC School system, Juvenile Repeat Offender Unit at WSPD, DSS, OJJ, and CenterPoint Human Services. JasonNet supports confidential case management information sharing across agencies and gives each agency update capability to only their own department's case information. This system will allow a more complete picture of patterns of individual juvenile violence to be identified early enough to assist with successful intervention strategies.

Community Safety Information System

The technology portion of SACSI in Winston-Salem involved the development of a GIS Application to meet our local research needs. For the development of this component, the initial needs assessment phase involved interviews with 53 individuals representing 19 organizations. This resulted in a summary of possible data sources and attributes, as well as a wish list of potential GIS applications that those interviewed would like to implement in Winston-Salem and Forsyth County. In some cases the initial interviews yielded somewhat inaccurate information concerning what was actually available electronically and the actual data sources. We also encountered different definitions of certain layers, such as neighborhood, which varies depending on the agency. (The FC Assessor's Office neighborhood is very different from the definitions of the City/County Planning Department, Police Neighborhood Watch, and Housing Neighborhood Development.) Additionally, quite a few of the application suggestions were not realistic for the original scope of the Initiative.

INDUS Corporation was given the task of developing a Community Safety Information System (CSIS) that would be customized to assist with the ongoing analysis required for the SACSI efforts in Winston-Salem. This was to be an Intranet system accessed via web browser using the Esri MapObjects and Internet Map Server products with Oracle Relational Data Base Management System (RDBMS) for the data repository. The major analysis tools chosen to be initially implemented in the CSIS were the following five basic applications:

A technical lead role was established locally in the WSPD to be a contact with INDUS and NIJ, as well as to work with local resources on the CSIS development. In October, 1998, personnel from NIJ and INDUS met with individuals from key city and county departments to give an update on CSIS , discuss data sources in the community and to request the initial data acquisition and regular updates that would be needed to make this project a success. Individual departments and agencies were also contacted by the local technical lead as the original list of data and sources was evaluated.

Data Layers to Support CSIS

These five analysis tools in the initial CSIS design would require over 30 layers of GIS data, with approximately twenty-five of them to be newly developed. We were fortunate to have a well-established GIS infrastructure in our community with existing coverages for the Cadastral (Parcels), Schools, FC municipalities, Fire Demand Zones (FDZs) , Hydrography, Police Beats, Police Sectors and a newly developed routable Street Centerline for address matching.

CHALLENGES WITH CSIS DEVELOPMENT

Data Layer Development

The success of the collaborative efforts to provide the necessary data for CSIS has been due in large part to an active Forsyth County/Winston-Salem GIS community and the people and hardware networking already established across departments and agencies. The city and county made the decision in 1993 to use the same GIS software (Esri) to facilitate data sharing. The history of collaboration also included jointly developing our FC Centerline http://itre.ncsu.edu/gis/projects/ws/ , a countywide Internet application hosted by the FC Assessor's Office http://maps.co.forsyth.nc.us and sharing city and county data for an ArcView public access application across multiple city/county departments. As the CSIS efforts began, a routeable Centerline for the entire county was just being finalized as part of TIIAP grant funding from the Department of Commerce for the development of the Integrated Network for Fire Operations (INFO) in Winston-Salem. http://www.ci.winston-salem.nc.us/fire/infoproj/index.htm

Some of our local challenges related to the application and database development have included: identification, acquisition, integration and standardization of data from multiple agencies, the maintenance of confidentiality and security, and geocoding. The new data layers to be developed for CSIS included abandoned and unfit housing, Alcohol Beverage Control (ABC) permit holders and violation locations, arrests, businesses, crime incidents and person involvement, WSPD nuisance violation locations, demographics by beat, demographics by school districts, public subsidized housing locations, student demographics and home locations and school districts. (See Appendix A.) Also, the existing school coverage was updated to add additional information from WSFC Schools for CSIS.

Business Layer

The business layers were developed using electronic sources of business locations and types of businesses that were available from three different sources. Business locations were available countywide from a WS Department of Transportation (DOT) study collected by summer interns in 1994. The businesses were categorized using Standard Industrial Classification (SIC) codes, a detailed indicator of land use. In addition to being four years out of date, the WSDOT information contained invalid addresses, misspellings and inaccuracies of SIC coding. Businesses were available citywide from the fire department using their very detailed property classification coding scheme and very accurate addresses. However, the data also included businesses that were no longer operational. The WS Business licenses from the Revenue system was an additional source of currently licensed businesses within city limits, with ordinance code and suffix allowing us to derive categories of businesses. However, this data had no address validation for the business location and even the position of the actual business location in an individual Revenue record could be in either the main record or one of possible multiple location annotation records. Processing the data to improve accuracy, correcting addresses and confirming validity was a local responsibility coordinated by the technical lead.

We initially looked at creating separate business layers by category, such as adult establishments, bars, car dealerships, convenience stores, hotels/motels, restaurants, etc. As INDUS personnel worked with us to develop the CSIS Oracle database design, the decision was made to maintain the businesses in the three original data source formats, with 3 different coding standards indicating the type of business. We knew that we did not have the local time and resources to reconcile and consolidate the three data sources into one during this initial phase. However, this DB design decision added an additional requirement for the development of queries to determine the business category, which is one issue that we are working to resolve with the developer.

Data Acquisition Issues

Data acquisition of certain layers has also presented some difficulties during the development of CSIS. In some cases, a trail of contacts was required before finding the right person who could assist us with acquiring certain data. We often encountered delays in receiving data and followup reminders were needed. Some of the delays were due to staff responsibilities within their own agencies. Due to school redistricting process, the CSIS data extraction by School for attendance, discipline, race/sex data was not received until mid July, 1999. Some agencies did not have the capability to extract their own data due to vendor specific applications or lack of technical expertise. For example, we were not able to get electronic crime data from the FC Sheriff's office. Therefore, there is no crime data for the county in CSIS at this time. In order to acquire the open cases for unfit and housing to be demolished, an on-site visit was required to the WS Housing Code Enforcement office to provide technical assistance with the development of a Crystal Report for data extraction.

Another issue that has added an additional level of complexity to the geocoding task for developing these CSIS layers is the different address standards, abbreviations, and misspellings in the original data sources. Addresses of students, home addresses of persons with offense involvement, ABC locations, Housing Authority project locations, business licenses required initial preprocessing cleanup and standardizations with Visual FoxPro programs before the data was sent to INDUS for migration into Oracle tables. This data cleanup was done at the WSPD by the technical lead.

Confidentiality issues have also been encountered, which have delayed and limited data which will be available for analysis. WSFC Schools made the decision not to release names of students or individual discipline and attendance records due to concerns about student record confidentiality. These school data confidentiality issues resulted in no school information identifiable by student. It was not until mid July, 1999, that we were finally able to acquire the aggregated discipline and attendance information by school, which has not yet been integrated into the application. Although, the Housing Authority of Winston-Salem (HAWS) was initially hesitant to share addresses of Section 8 housing, this has been resolved and these locations are available in the CSIS.

WSPD Transition to New Police Records System

In the midst of the development of CSIS, the WSPD implemented a new client server system on January 1, 1999, and migrated data from a Unisys mainframe to Oracle tables on a UNIX server with our new Pistol 2000 application. We also made a change from the Uniform Crime Reporting (UCR) standard to the Incident Based Reporting (IBR) codes on at the same time. To simplify the coding of the CSIS application, which would have had to use both UCR and IBR codes, depending on the date parameter, we agreed to preprocess our historical data to convert from UCR to IBR codes. At this point, the GIS application uses IBR crime category definitions, while the historical data is still in UCR code. This update will be completed with the next migration.

IMPLEMENTATION OF CSIS

The next section discusses how the Community Safety Information System (CSIS) is being implemented in Winston-Salem. INDUS Corporation developed this application using Microsoft Visual Basic 5.0 and 6.0, Seagate Crystal Reports 6.0, MapObjects 1.2, Internet Map Server (IMS) 2.0, (Environmental Systems Research Institute, Inc., Esri) and JavaScript 1.2 software. Data is stored in Oracle 7.33 RDBMS. It runs on a dedicated server, Dell dual processor, 450 Mhz, 2 - 18 GB drives, 500 MB RAM, which is housed at the WSPD. This application is password protected to limit access to only those privileged to see the sensitive juvenile data that it contains. Queries are submitted via html forms and converted by java script into SQL queries to Oracle tables. The results are passed to the web server where the map is created and sent as a graphics image to the client's browser. Some features of the application include:


CSIS APPLICATION


SACSI APPLICATION DISPLAY

NOTE: Due to the sensitive information in CSIS concerning juvenile crimes, this application is not available to the general public. The application is password protected. The URL shown on the screenshot is an Intranet address with limited access.

CSIS Crime Analysis Query


SACSI Crime analysis

Juvenile Crime Analysis


Juvenile Crime analysis

CSIS Classification of Demographics


Demographics

Winston-Salem/Forsyth County CSIS Data Repository

The data repository for the CSIS is collected from the contributing city county agencies and departments. Oracle 7.3.3 was chosen for the relational database because of our existing expertise and preference locally for this RDBMS . This repository consists of both spatial and relational data that require periodic updates with varying frequency, depending on the nature of the data. INDUS developed generic data migration tools and automated the geocoding process for CSIS. Each dataset source must be collected, cleaned with preprocessing, if necessary, then converted to a comma delimited ASCII file with quotes that is then imported into a file (dbf) to be geocoded with ArcView 3.1 in order to create the spatial data layers. Migration scripts have been developed to then load the appropriate Oracle tables with data that will support the querying capability of the CSIS.

Status of CSIS Application Development

The development of the CSIS has encountered delays and there are important issues that are still being resolved. The application is in the test phase and local training is scheduled for August 11, 1999. Some issues to be resolved include:

  • Applications for demographic analysis not fully developed as yet
  • Presentation output very limited, with no map output as yet
  • Printing requires screen print capture and pasting into Windows paint accessory
  • The ability to select business type was not initially implemented
  • Crime Type by Modus Operandi (MO), Time and Day does not support MO parameter or a specific Day of Week
  • Crime Type by Environmental factors not developed as yet
  • Crime by proximity doesn't enable user to compare incidents against larger area or allow proximity analysis with a selected layer
  • Insufficient edits in place to prevent users form submitting queries and other requests that are not supported

LESSONS LEARNED FROM CSIS DEVELOPMENT

  • The initial aggressive schedule was not realistic for all of the work on both ends. The technical kickoff meeting with requests for data from many departments was in October, 1998 and the final release of application was initially scheduled as December, 1998. The schedule was unrealistic and was not met.
  • There will be delays for development with leading edge technology due to the learning curve and new challenges faced.
  • There was not enough testing time built into the schedule before delivering the product.
  • Local expectations would have been managed better with more ongoing communication with the actual developers, and more on-site visits with preliminary demonstrations of the features being developed. This would have provided the feedback to keep the application on target for our needs. For example, the capability of querying by age and role parameters was not implemented in the first release of the application.
  • More awareness is needed to prevent the breakdown in communication between the Data Base design phase and the development and implementation. Modus Operandi (MO) was to be in the crime type query, but was dropped because this was not in the historical data from WSPD. This issue had been discussed with the DB design staff and tables with MO codes had been provided. WSPD data entered after 1/1/1999 in our new system will contain this information that was not in the historical data.
  • Expectations are also better managed through more formal documentation. A Functional Requirements document would have pinned down the functionality of the system and would have identified that queries were developed that supported this design. This would have prevented misunderstandings over how a proximity query functions, for example. The features documented in earlier presentations to the local SACSI have discrepancies from what has actually been developed.
  • A Test and Acceptance Plan and more developer testing before releases would have ensured more quality in the application releases. This also would have helped the local testing efforts to be more effective. Fixing one problem can introduce problems in other components, so testing of all prior combinations is also required.
  • When encountering technical issues that would require a design change or prevent a feature from being implemented, identifying and renegotiating these issues needed more involvement from the local SACSI technical contacts.
  • More time was spent on database design with not enough time spent ensuring that this design was then integrated with the query specifications. For example, the local expectation was the ability to perform analysis with specific business types. i.e. bars, convenience stores, car dealerships, hotel/motels, restaurants, churches. However, the application initially lacked a mechanism for selecting business categories with the predefined queries delivered.
  • The database design was focused more on providing generic loading capabilities for various data sources, rather than on supporting queries and reporting. This has led to delays in development of queries and problems with Crystal Reports. One problem that has caused delays is the storage of dates in character format, rather than date. Also, multiple joins to retrieve query results slow down response.
  • The version of Oracle used by developers should match the version used at the local site. The site version is Oracle 7.3.3, while the developers were using 7.3.2. Issues with date differences concerning two digit versus four digit year were found with these versions until a patch was applied to 7.3.2.

Local Challenges and CSIS Enhancements

We have many challenges facing us locally in order to maintain and enhance the CSIS. Crime related information has still not been provided for Forsyth County. We are still attempting to acquire this from the FC Sheriff's department. When and if this is received, a process to load this into CSIS and ensure that the existing queries work with this additional data will be needed. The WSPD historical crime data used in SACSI must be converted to IBR coding. INDUS will be updating the migration process scripts, which bring in the community data sources and create the GIS layers and update Oracle tables. Additionally, the street centerline network requires ongoing development and maintenance, as do the new layers developed for SACSI. The new demographic, discipline and attendance data just received from the school system remains to be integrated into CSIS, as well as the unfit and to be demolished houses. The new layer just developed at WSPD of the home addresses of habitual felons and repeat offenders would also be very appropriate to integrate into CSIS.

One important function we initially lacked was a way to identify specific types of businesses. i.e. show all of the hotels/motels, restaurants, car dealerships, bars, churches, etc. This is being resolved with the developer. Also, more local work is needed to investigate and consolidate valid businesses from the WSFD, WSDOT, and WS business license data. As land use codes are collected and coded in the cadastral coverage by the FC Assessor's Office in the near future, we will have another cross reference of the business location data that can be used as a cross check of our business coverages being enhanced. In order to improve the analytical functionality of CSIS, this data must be reviewed and updated.

As this application was initially implemented, we either were able to find crimes within proximity of a point or the crimes contained within a polygon coverage item to be selected from the map. Therefore, we were not able to find all crimes within 500' of school property, which is a polygon feature, nor were we able to identify crimes within a buffer area of a specific street. This limitation is being resolved with the developer. We also would like the capability of selecting all features in a layer and defining crimes either within the buffer defined of each feature, or the inverse - outside of the buffer, as in the street lights. The functionality of selecting a coverage and finding crimes within specific proximity of each item i.e. drug violations within 500' of all schools would then be supported.

Other issues to investigate are the slow response for some queries and the possibility of redesigning the Oracle tables and updating the Java scripts, Crystal Reports and SQL queries to convert dates from character attribute to date.

For the continued success of the CSIS application, the spatial and attribute accuracy of this information must be maintained. Providing analytical tools to researchers and allowing extraction of data for further manipulation is one goal. Maintaining data quality is another responsibility that will continue to take local time and resources and can not be compromised.



CONCLUSION

SACSI Initial Findings in Winston-Salem

According to the "SACSI Implementation Action Plan Initial Draft", June, 1999 , the key findings from the SACSI research analysis are that juvenile violence is clustered in four specific neighborhoods in Winston-Salem, with hot spots around certain blocks, dead-end streets and convenience stores. Older offenders often tutor younger ones, especially in drug violations. A small number of perpetrators are responsible for most serious crimes. The research is pointing out a need for after school and mentoring programs. The fundamental strategy recommendations from the Strategic Core team are that enforcement, intervention, and prevention activities are needed. Enforcement strategies will focus on repeat offenders and cleaning up problem locations. Intervention strategies will involve early identification and provide intervention for chronic juvenile offenders with priority treatment and services. The prevention component recommends a family-focused case management approach, with in-school and after-school activities in SACSI identified neighborhoods. Conflict resolution and anger management skill development with cultural sensitivity is also needed. Another issue identified by the SACSI analysis was that offenders 16 and 17 years-old had aged out of the WSPD juvenile repeat offender unit and had "fallen through the cracks", because they were not automatically passed on to the adult repeat offender unit. This has been remedied through new procedures within WSPD.

The SACSI project has provided a very much needed and greatly appreciated jump start to our community's endeavor to reduce juvenile crime. It has facilitated even more communication and cooperation across the local partnership of agencies. The CSIS, with this integration of web technology, is assisting us to make GIS more widely available. It is an excellent basic application that can be enhanced to meet the ongoing analytical needs of the SACS Initiative and has a lot of future potential for other cities.

Since Winston-Salem was the first CSIS site to be developed, we have had more technical difficulties with the VB/MO IMS solution technology. We still have lots of additional work ahead of us to fully utilize its potential. However, we are well under way in our community to deal with the critical issues concerning juvenile violence and to implement intervention strategies to redirect the lives of our children. The work being done in Winston-Salem will not only help our community, but will help to develop a model for other cities that will assist in crime reduction. We also have a responsibility with both the technology and in the intervention roles to continue this process and make a more proactive response to issues of juvenile violent crime in our community.


Appendix A

Data and Sources for the Initial Applications

Abandoned Housing

FC - Housing Code Enforcement

Abandoned Housing

WS - Housing Services

ABC Permit Holders

Alcohol Beverage Control (ABC) Commission

Accidents

WSPD

Adult establishments

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Arrests

WSPD PISTOL 2000 Arrest Module

Car dealerships

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Convenience stores

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Crime Incidents

WSPD PISTOL 2000 Incident/Investigation module

City Boundary

City/County Planning Department

Crime victims

WSPD

Demographic characteristics

WS Housing Neighborhood Development

*Drug markets - unavailable
(Locations for liquor/drug violations)

ABC Violation Locations
WS Abatement Violations

ABC Violation Locations

ABC Law Enforcement

WS Abatement Letters

WSPD

Fire Demand Zones

WS IS Department

Hotels/motels

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Nuisance Violations

WS - Housing Code Enforcement
FC - Housing Code Enforcement

Liquor houses

ABC Violation Locations
ABC Law Enforcement
WSPD Abatement Letter (Locations for liquor/drug violations)

Neighborhood Association Areas

City/County Planning Department

Nightclubs

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Property (Parcels)

FC Assessor's Office

Police Beats

WS IS Department

Public/Subsidized Housing

HAWS Housing Authority of Winston-Salem

Housing Authority Project locations and Units

HAWS Housing Authority of Winston-Salem

Restaurants

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

School Districts
elementary, middle, and high schools

City/County Planning Department

Schools

FC School System IS
City/County Planning Department

Sectors

WS IS

Street Centerline

WSPD initially, then FC Assessor's Office

Street Lights

WS DOT

Students

FC School System IS

School Attendance

FC School System IS

School Discipline

FC School System IS

Person Data

WSPD PISTOL 2000 Name Module

Youthful offenders

WSPD PISTOL 2000 Name Module

Banks

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Churches

WS Business License
Employer Data 1994 WS DOT
WS Fire Department

Railroads

FC Assessor's Office

Hydrology

FC Assessor's Office

County/City

Esri - County.shp. For US

FC Corporate Limits

City/County Planning Department

Major roads

Centerline attribute 'Road_Class'

Arrestees

WSPD PISTOL 2000 Arrest Module




Author Information

Julia B. Conley
Senior Systems Analyst
Winston Salem Police Department
Information Systems Division
City of Winston-Salem
725 N Cherry St
Winston-Salem, NC 27102
(336) 773-7864 FAX (336) 773-7996
juliac@wspd.org



Elizabeth Groff
National Institute of Justice's
Crime Mapping Research Center
810 7th Street N.W.
Washington, D.C. 20531
(202) 305-3301 FAX (202) 616-0275
groffe@ojp.usdoj.gov



Tim Lesser
IS Senior Analyst
Department of Information Systems
City of Winston-Salem
P. O. Box 2511
Winston-Salem, NC 27102
(336) 727-2848 ext 3439 FAX (336)727-2874
timl@ci.winston-salem.nc.us