Mechanism of the adaptive safety analysis and monitoring system

The problem of detecting,tracking,and counteracting terrorist networks was resolved by the adaptive safety analysis and monitoring system based on hidden Markov models and dynamic Bayesian networks for assisting intelligence analysts to identify terrorist threats and predict possible terrorist actions. The basic premise was that terrorist networks can be evaluated using transaction based models.Analogous to the target tracking problem where states(location,velocity,etc.)are observed through noisy measurements, the true states of terrorist activities are detected against a background of noise transactions data(people, places,things)that appear to be suspicious by using hidden Markov models based on forward variable and log likelihood ratio algorithm and by providing local assessments as outputs that are transformed into soft evidence.Bayesian networks were maintained by higher level(decision making)agencies functioning as fusion centers,and by pooling the summarized information in the form of soft evidence to support belief updating. Combined with the counter-terrorism network models,the adaptive safety analysis and monitoring system integrates,shares and identifies information,in order to evaluates and predicts the terrorist network states. It can thus provide early warnings to facilitate preemption and/or support strategic decision making.