Crime forecasting is a new approach for crime prevention, using statistical tools to predict where crimes will take place and sending police to those locations before the crimes occur. It uses data about past crimes and some knowledge of criminal behavior to project possibilities on a computer-generated crime map, and is based on the idea that criminals tend to be habitual.
Computer-based forecasting can work with more variables than the human brain can easily handle, integrating seasonal patterns, annual patterns, and emerging trends to be 80 percent accurate, according to researchers at Carnegie Mellon University.
The trend analysis software used is similar to that used by big retailers and forest rangers. The Justice Department’s Project CeaseFire experiment uses crime forecasting combined with outdoor audio sensors that detect gunfire; it automatically telephones residences and businesses in the vicinity of the noise to give witnesses the opportunity to report information confidentially. However, the systems are by no means perfect.
Abstracted by the National Law Enforcement and Corrections Technology Center(NLECTC) from the Popular Mechanics (07/04) Vol. 181, No. 7, P. 36; Wilson, Jim .