Event forecasting system

A predictive policing system for use over a telecommunication system including a global system of computer networks:

• a network interface operable to be placed in communication with a telecommunication system including a global system of computer networks;
• at least one crime prediction processing device constructed to be placed in communication with the network interface and to carry out a set of one or more instructions to:
• access at least one historical crime report over the telecommunication system, the historical crime report include a set of crime data corresponding to a plurality of criminal events, the crime data including a crime location, a crime time, a crime date, and a crime type associated with each criminal event;
• receive at least one real-time crime report over the telecommunication system, the real-time crime report including a second set of crime data corresponding to at least one criminal event having taken place within a selected time interval, the second set of crime data including a crime location, a crime time, a crime date, and a crime type associated with each criminal event;
• determine a spatial domain based on a set of selected crime locations received as part of the crime data;
• subdivide the spatial domain into a plurality discrete geographic regions;
• process the crime data with a crime forecasting algorithm using a parametric equation including an iterative Expectation-Maximization formula that takes into account both spontaneous crime events and triggered crime events to produce a set of one or more discretized probabilities for a particular crime type;
• rank the set of discretized probabilities in accordance with a display threshold criteria;
• for one or more discretized probabilities, output a forecast assigning a geographic region to at least one crime type;
• combine the forecast with a visual map element depicting the geographic regions with at least one discretized probability meeting the display threshold criteria to create a predictive crime report having at least one geographic region and a probability of a future crime occurrence at a particular time interval associated with at least one crime type;
• transmit the crime report over the telecommunication system for display on a remote display screen for use in deploying policing resources to one or more high risk areas as shown in the predictive crime report;
• update the forecast for each new set of real-time crime report received.

A method of forecasting crime comprising the steps of:

• providing a processor with a memory unit containing a set of processor instructions;
• providing a network interface constructed to communicate with the processor and a remote source over a large scale network;
• receiving a plurality of historical crime reports from the remote source, each crime report including a set of crime related data including at least crime location, crime time, crime date, and crime type for each crime in each historical crime report;
• following the processor instructions to process the crime related data to assign one or more discretized probabilities of a crime type occurrence and a crime type location with each crime type using a parametric equation including an iterative Expectation-Maximization formula taking into account spontaneous events and triggered events and rank the set of discretized probabilities in accordance with a threshold criteria;
• receiving a digital map of a region of interest from the remote source over the large scale network;
• and
• following the processor instructions to transmit a display to one or more remote sources in communication with the large scale network, the display including at least a portion of the digital map and one or more delineated areas wherein the likelihood of a crime of a particular type occurring are more probable than not and the associated discretized probability meets the threshold criteria.