Using case-base data to learn adaptation knowledge for design

摘要: One advantage of Case-Based Reasoning (CBR) is the relative ease constructing and maintaining CBR systems, especially as a number commercial tools are available. However, there areas that current have not yet addressed. these easing or automating acquisition adaptation knowledge. Since tasks like design planning typically require significant amount adaptation, systems for still do fully benefit from CBR's promise reducing development effort. To address this, we developed several "knowledge-light" methods learning knowledge cases in case-base. These perform substitutional both nominal numerical values, suitable decomposable problems, particular formulation configuration. Tests performed on tablet domain show promising results. The automatic present can easily be incorporated general-purpose tools, thus further contributing to cost systems.