Case-Based Reasoning Approach

As stated in the previous part, design activity is based on a generic problem solving process which begins with problem definition and description, involves various forms of analysis, might include simulation and modelling, moves to generation of solutions and thence to the evaluation of alternative solutions to the problem. Decisions characterise every stage of this process. The process takes place across many levels of abstractions and is iterative in form. Meanwhile, practice shows that often it is more efficient to solve a new problem by starting with a solution of a previous, similar problem than to generate the entire solution from scratch. Whenever it is easier or more convenient to reuse experience, humans prefer to do so rather than to derive a completely novel solution. A design engineer may find it difficult to determine the similar elements between a new problem and massive historical data; moreover, the similarities are often not easily noticeable. In order to facilitate the design process and to reduce the required development time, a decision supporting system utilising the case-based reasoning approach could be used to support process and product design.

Case-based decision support systems try to help in decision making by providing cases similar to an actual problem situation. The cases are usually advice or a useful piece of information. In the context of decision support, the problem is a situation description in which information is missing. The goal is to complete the description of the situation until the demand for information is satisfied. Thus, the problem solving process is to provide the user with records to satisfy his demands appropriately. This chapter reviews the different models of the case-based reasoning and describes the case-based methodology to support decision making in design.