The Prisma System: intelligent agents working on crime pattern analysis supported by geographic information systems

Abstract

The process of extracting useful knowledge from large databases became one of the tasks of prior importance in today’s organizations. The collection of excessive amount of information makes very difficult its treatment and analysis without appropriated means. Police Departments are real examples of organizations that currently debate themselves with situations involving large volumes of distributed information and requiring effective real time decision making. Some of these situations are critical in the normal Police Department’s activities, namely the ones related to Crime Pattern Analysis. These are concerned with the recognition of spatial and temporal regularities in reported crime and the ability of predict future criminal activity. This is very important due the possibility to provide effective elements to increase patrol actions, improve priority investigations or even perform better public notification. Through the combination of Multi-Agent Systems and Geographic Information Systems technologies we design a computational system Intelligent Crime Pattern Analysis: the Prisma system. It considers a community of intelligent agents, divided essentially into two classes, that will be responsible respectively to populated specialized Data Marts and make Criminal Patterns Identification. With Prisma, Police Departments will be able to examine patterns related to notified incidents and analyze their movement in relation to police initiatives.