Identifying Critical Dependencies Using General Bayesian Network: An Application to Aviation Operation

This paper proposes the general Bayesian network as a tool for identifying critical dependencies among complex human beliefs, especially for the case of airline service industry. When we consider an overall efficacy of the flight operation, to understand underlying beliefs between pilots and controllers is critical. However, even for the experts in organizational behavior, it is difficult to identify dependencies among such the beliefs. To tackle this ambiguous issue, we suggest the use of the general Bayesian network to automatically generate interesting hypotheses. The experimental results reveal the possibility of selective attention to the critical dependencies for achieving enhanced efficacy.