Developing a Crime Hotspot Prediction System for the Nigeria Police Force in Lagos

Predictive policing based on crime hotspot analysis can improve the efficiency of police resource deployment in Lagos, the city with the highest crime incidence in Nigeria. This study developed a crime hotspot prediction system for the Nigeria Police Force Lagos State Command. A professional system design methodology was applied, drawing on structured consultations with 14 senior NPF officers and 8 crime analytics specialists, review of three years of anonymised crime incident reports from Lagos State Police Command, and benchmarking against crime analytics systems deployed by the Metropolitan Police, Kenya National Police Service, and SAPS. The system designed specifies a geospatial data pipeline ingesting incident reports, infrastructure data, and environmental variables, a kernel density estimation layer for baseline hotspot identification, a short-term predictive component using random forest trained on temporal crime patterns, and a patrol dispatch recommendation module. Data anonymisation and civil liberties safeguards are embedded in the framework design in accordance with Nigerian constitutional provisions. Alert generation protocols for shift commanders are specified. Expert review by ten crime analytics and policing specialists confirmed the system's operational utility. The study recommends piloting the system in Lagos Island and Surulere Divisions where historical data quality was highest, establishing a crime analytics unit within Lagos State Police Command to manage the system, and conducting quarterly bias audits to detect disparate impact on specific communities.

Keywords: crime hotspot prediction, predictive policing, Nigeria Police Force, Lagos, geospatial analytics