Stephen Hollifield, CIO City of Richmond Police, Virginia Visual Predictive Analytics in Law Enforcement
Crime in Richmond
Data Rich yet Information Poor (~50 million records) No way to compare current crime with vast wealth of historical data Statistical forecasting limited to linear perspectives Analysis only available to limited personnel with delayed and cumbersome distribution Our Challenge
Find: Hidden patterns & relationships Analyze: Predict likelihood of violent crime Automate: Updated forecasts every 4 hours Distribute: Enterprise-wide access Leverage: Deploy resources effectively Our Strategy
Data –Bring data from multiple sources into data warehouse –Discover hidden patterns and relationships Enabling / Interfering Factors Weather Moon Phases Triggers City Events Paydays Holidays Day of Week Time of Day Crime Records Data Sources
Predictive Results
Crime Maps
Visual Discovery
KPI - Incident Analysis
The Cost
Homicide -32%-40% Rape -19%-8% Robbery -3%-20% Aggravated Assault -17%-5% Burglary -18%-7% Motor Vehicle Theft -28%-26% Statistics
Have we been able to reduce crime? Do people feel safer? Are certain crimes occurring less often? Are we making efficient use of our resources? Measuring Success
SMTWHFS 7 & 30 Day Analysis –Predict intensity of crime by 4 hour windows within 7 and 30 day forecasts –Provide single click interface directly to GIS perspective for each 4 hour window –Provide “what if” scenario options based on deployment tactics SMTWHFS Future
Stephen Hollifield, CIO City of Richmond Police, Virginia Visual Predictive Analytics in Law Enforcement