Author :
Singh, Satnam ; Tu, Haiying ; Allanach, Jeffrey ; Areta, Javier ; Willett, Peter ; Pattipati, Krishna
Author_Institution :
Electr. & Comput. Eng., Connecticut Univ., CT, USA
Abstract :
This paper discusses the prior information of choosing model parameters for constant threat from well planned, sophisticated and coordinated terrorist operations against geographically diverse targets, which causes significant loss to human life and property. A semiautomated model-based tool to detect and track terrorist activity are developed based on the analysis of prior terrorist attacks through the clues about the enabling events and the information from open sources. A pattern of transactions is a potential realization and prediction of a possible terrorist event. The HMMs is a stochastic model used to detect the monitored terrorist activity and measure local threat levels. The Bayesian network probabilistic model is well-suited for modelling the global threats and for computing/assessing the overall threat. Optimization techniques can be used to allocate counter terrorism resources and software to track multiple terrorist activities using multitarget tracking algorithms for intelligence analysis.
Keywords :
belief networks; hidden Markov models; optimisation; target tracking; terrorism; Bayesian network probabilistic model; HMM; counter terrorism resource; enabling terrorist event; event prediction; geographically diverse target; global threat modelling; human life loss; intelligence analysis; local threat level measurement; model parameter; multitarget tracking algorithm; optimization technique; property loss; stochastic model; terrorist activity monitoring; terrorist attack; terrorist operation; threat computing; threat modeling; track multiple terrorist activity; track terrorist activity detection; transaction pattern; Bayesian methods; Computer networks; Event detection; Hidden Markov models; Humans; Information analysis; Monitoring; Stochastic processes; Target tracking; Terrorism;
Journal_Title :
Potentials, IEEE
DOI :
10.1109/MP.2004.1341780