Getting to Know the Crime Analysis Solution

Crime analysis is an essential law enforcement function used to enhance public safety, identify emerging trends, allocate resources, and plan crime-prevention strategies. 

Successful data-driven strategies use the work of crime analysts to reduce crime by focusing resources on high-crime places, high-risk offenders, and repeat victims.

Denise O'Donnell, Former Director of the Bureau of Justice Assistance

Modern law enforcement agencies depend upon the work of crime analysts, helping everyone from beat officers to command staff use analytical products to answer questions and make more effective, efficient, and just decisions. To support the needs of their users, effective crime analysts must learn a wide array of analytical methods and have access to data and tools necessary to perform those analyses correctly. Before analysis can even be performed each day, those analysts must ingest data from multiple data sources, fix a variety of data quality problems, and map the data. 

As a result, some organizations struggle with maximizing the impact of their crime analysis capability. Crime analysts often lack easy to learn tools that help them perform the wide set of analytical methods they are asked to perform and are forced to perform tedious manual data import processes due to lack of automation tools.

The Crime Analysis solution improves productivity of crime analysts with specialized tools and a tailored desktop GIS interface for mapping data and performing spatial analysis, making it easier to quickly generate high-quality crime analysis map products. Improved crime analyst productivity helps law enforcement decision-makers optimize resource allocation, solve cases, and implement more effective crime reduction strategies. Crime Analysis is typically implemented by law enforcement agencies that want to use data-driven strategies to reduce crime and improve police operations.

To perform analysis, crime analysts must first import data from disparate record systems and prepare it for use in GIS. Data import and management activities like geocoding, date field transformations, and appending of spatial data must be performed daily. These time-consuming but necessary activities leave little time for actual analysis, limiting analyst productivity. Crime Analysis provides tools to help analysts automate data import and simplify common data management activities.

The Update Features from Incident Records tool can be used to import data from a spreadsheet, CSV, or database (or  Microsoft Access ) table, map records using addresses or x,y-coordinates, and update an existing local or web layer. The tool can be automated to run monthly, weekly, daily, or hourly using ArcGIS Pro scheduling tools, dramatically reducing the amount of time analysts spend on data preparation.

Once data has been imported into ArcGIS Pro, Crime Analysis addresses common data management needs with tools for enhancing the attributes of your data. The solution includes tools to help convert dates stored as text; use a date field to create new date part fields like day of week, hour, month or year; append the police beat or division to an incident record, or enrich boundary data with population or other socioeconomic data from Esri’s Living Atlas.

Analysts occasionally annotate their maps with descriptive text boxes, images, arrows, or other shapes. The Add Graphics Layer tool allows analysts to create and store this content, and a graphics layer can be toggled on or off in the Contents pane like any other layer. The CrimeAnalysisSolution ArcGIS Pro project included with the solution has example maps of each tool output and uses graphics layers to provide explanatory text to each map.

Crime analysts ask a wide variety of analytical questions on behalf of the organizations they support, requiring them to query their data from a variety of date-based, spatial, and field attribute perspectives that can be challenging to perform without the use of GIS. Crime Analysis provides selection tools that can be used individually or in tandem to interrogate incident data and focus on specific problem subsets.

In addition to the standard selection tools available in ArcGIS Pro, Crime Analysis includes Select Layer By Date and Time, a tool for selecting records based on date and time ranges or date parts, for example by day of week, by month, year, hour, or dynamic time ranges, for example, last 30 days. When combined with attribute-based queries and location-based queries, analysts have a complete set of rich selection tools with which to build complex composite queries of their data.

Crime analysts support the data-driven crime reduction efforts of law enforcement agencies by using tactical analysis methods to identify short-term crime patterns, and strategic analysis methods to support the identification and evaluation of long-term crime problems and trends. Central to these methods are spatial analysis, specifically hotspot identification, but analysts can be challenged with understanding which techniques are available and are appropriate to the type of analysis they are trying to perform. Crime Analysis curates new and traditional spatial analysis tools appropriate for tactical and strategic crime analysis, making it easy for analysts to discover the right tool for the job.

Crime Analysis includes a group of analysis tools that can be used in tactical analysis to identify short-term crime patterns. 80-20 Analysis is used by crime analysts to identify the small number of locations that are disproportionately responsible for a majority of incidents, for example, 80 percent of calls for service come from only 20 percent of locations. By identifying these repeat problem locations, law enforcement agencies can reduce overall incident volumes by creating specific strategies to handle those locations that are disproportionately responsible for the problem. Summarize Incident Count helps analysts counts the number of incidents in each polygon or line, symbolizing the counts using a graduated color ramp. Crime analysts typically create graduated color maps by incident count to contrast the frequency of a problem across police or neighborhood boundaries. Graduated color maps of incidents counted by street segment are also a common method for identifying micro-places where crime and disorder problems cluster. Crime analysts use Summarize Percent Change to support CompStat or other types of crime control processes in which decision makers need to understand how a count of incidents in the current period (year to date, last 28 days, etc.) compares to the previous period across each beat or district.

Kernel Density is the tool most frequently used by crime analysts to identify hotspots and is available in the Crime Analysis solution for ArcGIS Pro users with a Spatial Analyst extension license. The solution also includes the Minus tool for evaluating change in density between two periods of time. Analysts use this method to evaluate impact of police operations on hotspots and well as to explore how hotspot patterns change over time. Density-based Clustering is a new tool that uses machine learning algorithms for finding incident clusters and can act as a complementary or alternative to Kernel Density.

Crime Analysis also supports the identification, classification, and prediction of repeat and near repeat victimization patterns in your data.  Repeat Victimization  is an empirically observed phenomenon where when a location becomes the victim of a specific type of crime, that location has an elevated risk of subsequent victimization, a risk which drops precipitously over time. A related phenomena is known as  Near-Repeat Victimization , whereby locations immediately around a location that has been victimized are at elevated risk, a level that drops both as time passes, and as the distance from the originating victim increases. Repeat and Near Repeat patterns have been observed across a wide variety of crime types, are believed to be spatial artifacts of serial offending behaviors, and thus have important implications for crime prevention and investigations.

Crime Analysis includes tools to support analyses of Repeat and Near Repeat Victimization patterns. The Export Near Repeat Calculator Table tool can be used with Temple University’s  Near Repeat Calculator  to identify statistically significant patterns of Repeat and Near Repeat Victimization. The Near Repeat Classification tool can be used to identify historical patterns of repeat and near-repeat victimization, which can support investigative efforts to connect and clear cases that were not otherwise known to be part of a pattern. The Calculate Prediction Zones tool can be used to generate daily tactical crime risk predictions based on Repeat Victimization and Near-Repeat Victimization risk for recent crime locations. These risk predictions can be used to inform police operational strategies for crime reduction.

Lastly, the solution includes several tools used in strategic analyses of larger datasets to understand long-term problems and trends. Optimized Hotspot uses the Getis-Ord Gi* method to identify statistically significant hot and cold spots of incidents. The Create Space Time Cube By Aggregating Points and Emerging Hotspot extend the Getis-Ord Gi* method, organizing incidents into geographic bins sliced by time steps, for example, month or year, and applying the statistical method to not only identify statistical hot and cold spots, but also characterize how they change over time. This allows commanders to optimize their response with strategies and tactics appropriate for the unique spatiotemporal character of each hotspot. Colocation Analysis uses the colocation quotient statistic to test whether the distribution of two subsets of data, for example burglaries and pawnshops, crashes and crimes, or field stops and violent crime, are more frequently near each other than would be expected given the distribution of data over the overall jurisdiction. Understanding spatial colocation between two datasets can help analysts to better understand factors influencing a crime problem and to evaluate changes after responses.

Crime analysts support investigations by using data to explore the relationships between people, events, and locations. Investigative analysis helps investigators uncover criminal networks and understand suspect activity patterns through analysis of cell phone records, financial transactions, and other investigative data sources. These forms of analysis have previously required specialized, expensive, single-purpose software programs, but the Crime Analysis solution for ArcGIS Pro supports all of these workflows within a single program.

The Points to Track Segments and Generate Origin-Destination Links tools support common investigative workflows such as visualizing summary financial transaction data to identify patterns of money laundering and fraud, temporal sequencing of GPS tracks or incident data, and visualizing spatial relationships such as between locations of motor vehicle thefts with where those vehicles are recovered. 

Additionally, Crime Analysis includes a rich set of tools to support the geographic visualization and analysis of cell phone records. The solution includes tools to help analysts map cell sites and sectors, associate legally obtained call detail records to those locations, summarize calls by location, visualize suspect calls by time and space, explore spatio-temporal relationships between call records of investigative targets, and prepare maps for presentation in court.

To share their work with stakeholders and decision makers, crime analysts prepare information products summarizing their results. Analysts get requests to deliver these products in both hardcopy and interactive forms. To support the needs of their customers, analysts need flexible options to deliver the results of their analysis in a wide array of forms.

Crime Analysis allow users to make standard hardcopy map layouts as well as tabular reports suitable for printing or sharing as a PDF.

The solution also includes a rich library of interactive charts, including a Calendar Heat Chart to visualize temporal concentrations by hour of day and day of week. Additionally, Crime Analysis allows users to build their own Link Charts with tools for link analysis. Users can define entities (persons, places, events) in your layers and non-spatial tables, then define relationships to connect the layers and tables together. Link Charts can be visualized clustered, hierarchically, or organically, and users can make selections in the link chart that are connected to the underlying layers. Entities and relationships in a link chart that involve spatial data can also be visualized on a map. Link analysis methods like path analysis, centrality analysis, clustering analysis, and neighborhood analysis are also available, and any link chart can be exported as a PNG, GraphML, or XPS file.

For fully interactive information products, Crime Analysis can be used to publish layers and maps into an ArcGIS Enterprise or ArcGIS Online organizational portal to make dashboards, story maps, or other interactive web mapping applications.

The Crime Analysis solution delivers a set of capabilities to help you manage incident data, conduct tactical and strategic analysis, perform investigative analysis and share web-based or hardcopy information products with decision-makers.

This solution can be deployed into your ArcGIS Online organization using the ArcGIS  Solutions app . The solution is fully supported by Esri and includes  documentation  on how to use the apps in the solution as well as load your data and configure the solution to meet your organizational needs.

ArcGIS Solutions for Public Safety help police, fire, and emergency management agencies improve operations and enhance services provided to the public. Visit  ArcGIS Solutions  for more information.