A cognitive system for adaptive decision making

This paper seeks to investigate how the existing knowledge in a cognitive system can be used to help fill-in an incomplete situation picture. This is motivated by the human brain´s innate ability to use new incoming information together with stored knowledge to fill in gaps of the whole picture in the mind. The research focus of this paper is on the inference or retrieval of information that is relevant to the current circumstances, but not directly obtainable from actual observations. The objective is to derive methods to help uncover as complete a situation picture as possible with high degree of confidence. Two computational approaches to enhance situation awareness for adaptive decision making given partial information are proposed. The proposed approaches enhanced an existing system Dynamic Bayesian Reasoning and Advanced Intelligent Network (D´Brain) which is a cognitive based dynamic reasoning system that uses Bayesian network as the underlying knowledge representation. Simulations are carried out to test the feasibility of the proposed approaches. The results show that they are promising approaches to fulfil the objectives in enhancing situation awareness for adaptive decision making.