Automating concept development by analogical reasoning for advanced decision aids

Compared with conventional knowledge-based decision aids, decision support systems for product concept design must support concept development at the initial stage of human decision making. To support such processes, we must implement both interactive support and automated concept development. This paper is concerned with the latter, and proposes an automated concept development method using analogical reasoning. The features of the method are that it generates a new product similar to a given product with unique concepts and that it augments domain theory to expand the applicability of the new product. To validate the performance of the proposed method, we have applied the method to an electric appliance domain. From our experiments, the method has developed not only well-known concepts but also some unique concepts which have not been produced. Although the proposed method is very simple, it has a highly extensible architecture. Therefore, we conclude that the system can be used to support the initial stage of human decision making.