Automatic discovery of criminal escape points in urban areas: a genetic approach coupled with a simulation model based on collective intelligence

• Adriano Melo UNIFOR
• Vasco Furtado UNIFOR
• André L.V. Coelho UNIFOR

Abstract

Simulation of criminal activities in urban environments is an asset to decision makers on the police force. In order to perform preventive actions, the police needs to understand the behavior of criminals and their response to possible actions or configuration of patrols. However, simulation tuning is not trivial especially when it is presented as an optimization problem. In this paper, we describe a solution for the allocation of criminals into gateways (start points of criminals in the simulation) via the use of Genetic Algorithms (GAs). The use of GAs allows for the discovery, in an automatic way, of gateway configurations that, when used in the simulation, produce crime in the same distribution of real data. In the paper, we also show that, by making use of background knowledge, the system can provide suggestions of gateway configurations that are more plausible from the police expert viewpoint.

References

Calvez, B. e Hutzler, G. (2005), Automatic tuning of agent-based models using genetic algorithms, Procs. of the 6th International Workshop on Multi-Agent-Based Simulation (MABS), pp. 41-57.

Cansado, T. M. L. (2005), Alocação e despacho de recursos para combate à criminalidade. Dissertação de mestrado, UFMG, Belo Horizonte.

Bonabeau, E., Dorigo, M. e Theraulaz, G. (1999), Swarm Intelligence: From Natural to Artificial Systems. Santa Fe Institute Studies in the Sciences of Complexity Series. Oxford Press.

Buchanan, M. (2002), Nexus: Small Worlds and the Groundbreaking Science of Networks. W. W. Norton, New York.

Bussab, W. O. e Morettin, P. A. (2003), Estatística Básica, Saraiva.

Furtado, V., Melo, A., Coelho, A. L. V. e Menezes, R. (2007), A crime simulation model based on social networks and swarm intelligence, Procs. of 22nd ACM Symposium on Applied Computing (SAC 2007), pp. 56-57, Korea.

Furtado, V. e Vasconcelos E. (2006), A multiagent simulator for teaching police allocation. AI magazine 27(3), pp. 63-74.

Kaza, S., Xu J. and Marshall B. e Chen H. (2005), Topological analysis of criminal activity networks in multiple jurisdictions, Procs. of the National Conference on Digital Government Research, ACM International Conference Proceeding Series, Vol. 89, pp. 251-252.

Liang, J., Liu, L. e Eck, J. (2001), Simulating crimes and crime patterns using cellular automata and GIS. Department of Geography and Division of Criminal Justice. University of Cincinnati, Cincinnati.

Mitchell, M. (1996), An Introduction to Genetic Algorithms, MIT Press.

Pentico, D. W. (2007), Assignment problems: A golden anniversary survey. European Journal of Operational Research 176(2): 774-793.

Reis, D., Melo, A., Coelho, A. L. V. e Furtado, V. (2006), GAPatrol: An evolutionary multiagent approach for the automatic definition of hotspots and patrol routes, in: J. S. Sichman; H. Coelho; S. Oliveira (Eds.), Procs. of IBERAMIA/SBIA 2006, Lecture Notes in Artificial Intelligence (LNAI) 4140, pp. 118-127.

Xue, Y. and Brown, D. (2006), Spatial analyses with preference specification of latent decision makers for criminal event prediction. Decision Support Systems, 41(3), pp 560-573.

Weiss, G. (1999), Multiagent Systems: A Modern Approach to Distributed Artificial Intelligence, MIT Press.

Zipf, G. K. (1949), Human Behaviour and the Principle of Least-Effort, Addison-Wesley, Cambridge MA.