Risk evaluation using evidence reasoning theory

The evidence reasoning theory can generally represent uncertainty associated with propositions and datum and clearly distinguishes probability uncertainty caused by information deficiency from unknown caused by zero information. This paper proposed the conception of the risk basic strength assignment and built the general method and model of risk evaluation based on the evidence reasoning theory. This method permits experts to evaluate the risk event completely on their own professional experiences and knowledge independently. The new risk integrated basic strength assignment and the new risk factors integrated basic belief assignment generated by the Dempster-Shafer rule of combination both synthesize several experts' free judgments and reserve experts' characteristic cognition at the same time. Thus, experts are released from the traditional evaluation method which restricts them to assign probability to certain risk factors under certain risk occurring model. By this method we can get the integrated representation of the experts' knowledge and extract the experts' knowledge accurately.