Modelling Case-Based Reasoning in Situation-Aware Disaster Management
Situation-aware disaster management (SADIM) is a continuous decision-making and action taking process. This process requires prior knowledge of the ongoing environment and context. SADIM comprises two major processes: situation awareness (SA) - a cognitive process that assesses current situations and anticipates future situations in the environment; and, disaster management (DM) which is a decision-making process of preventing, preparing, responding, and recovering for and from a disaster. One of the decision-making technologies used in current SADIM is case-based reasoning (CBR). CBR is used for the disaster management process only. Situation awareness process in current SADIM is carried out using domain rules, statistical reasoning and other methods. This paper therefore presents a method of using CBR to carry out both situation assessment and disaster management decision-making processes in SADIM, building on previous work focusing in SA alone. Using CBR for both processes provides the capability of using past experiences to understand the state of the environment and also solve specific disaster problems. The paper evaluates the method through implementation in disaster prevention in the petroleum drilling domain for early kick detection to prevent a blowout disaster. The results show a clear improvement in similarity assessment and problem-solving prediction to prevent disaster.