Retrieving Reusable Design Cases: Exploiting Adaptation Knowledge in Design Reuse

Abstract

Case-based reasoning has been identified as a potentially fruitful candidate technology with which to investigate the development of automated design systems. Two critical stages in case-based design are design retrieval and design adaptation. In the former, designs that can be reused for a new design problem must be located. In the latter, retrieved designs must be modified to meet the specific demands of the new target situation. In this paper we will address both of these stages in the context of a case-based software design system called Deja Vu. In particular, it will be argued that the notions of design reusability and adaptability are intricately linked and an approach will be described which allows the adaptation requirements of design cases to be accurately predicted during retrieval and subsequently exploited during adaptation. We argue that this approach benefits from improved retrieval accuracy, exibility, and greater overall problem solving effcacy.