An artificial intelligence (AI)-driven crime forecasting tool developed by researchers at 最新天美传媒 is showing promising results in Sarasota, where it was associated with a nearly 20% reduction in crime. The tool uses AI and criminology-based forecasting to help law enforcement target identified "hotspots."
This interdisciplinary project brought together faculty from the Department of Criminology and the Bellini College of Artificial Intelligence, Cybersecurity and Computing, who used ten years of historical crime data to develop the advanced forecasting models that predict the geographic locations where crimes are likely to occur. Forecasts were then shared with leadership in the Sarasota Police Department and used to proactively deploy police officers during the month-long pilot period.
According to the research team鈥檚 citywide analysis, Sarasota experienced a 19.6% reduction in total crime after accounting for longer-term trends and seasonal patterns. Areas receiving targeted law enforcement deployments showed the largest decreases, and researchers also found evidence that crime reductions extended into nearby surrounding areas.
"The pilot appears to have produced benefits not only in the targeted locations but also in adjacent areas, rather than simply displacing crime elsewhere," said YongJei Lee, PhD, assistant professor in the Department of Criminology.
The project, which includes SooHyun O, PhD, Seungbae Kim, PhD, and graduate student Abel Le, along with Lee, highlights the growing role of interdisciplinary collaboration in addressing complex community challenges. By combining criminology, artificial intelligence and applied public safety partnerships, researchers hope to create more effective tools for crime prevention and resource allocation.
Future phases of the project will include additional model refinement, continued collaboration with the Sarasota Police Department, another field experiment planned for Fall 2026, and expansion into juvenile crime location forecasting. Researchers are developing a data-sharing system and exploring the creation of an application to streamline forecasting analyses and predictions. The team is also working on the next generation of the model, called AlphaGeo, which would incorporate Google Street View and satellite imagery to further improve forecasting accuracy.