A consensus support system for deciding a combination of risk-reducing plans under uncertain parameters

In order to avoid risks, experts design risk-reducing plans, set attribute values as parameters on risk-reducing plans (e.g. effect, cost, and so on), and decide an appropriate combination of the risk-reducing plans. The combination is decided using the objective function and constraints consisting of the parameters. However values of the parameters vary among experts, experts select and adjust a certain parameter until reaching their consensus on the combination. In this process, experts often select a parameter that can not make combinations agreed if its value is adjusted. Also this calculation of generating one combination takes much time because this combinatorial problem is large-scale. In order to solve two kinds of problems, we propose the consensus support system for deciding a combination that has two functions: one is presenting parameters to be adjusted using the gaps of the parameter preferences and the magnitudes of the parameter preferences, and the other is the high-speed solver of the combinatorial problem by applying Branch and Bound method. We implement our consensus support system and evaluate the effectiveness of the system. The proposed system can present the parameters that can lead to the agreed combination with 60% accuracy. And the system can derive the combination in about 20 seconds, while the existing method takes about 1800 seconds for deriving the combination.