Adaptation-guided retrieval: questioning the similarity assumption in reasoning

摘要: Abstract One of the major assumptions in Artificial Intelligence is that similar experiences can guide future reasoning, problem solving and learning; what we will call, similarity assumption. The assumption used reasoning systems when target problems are dealt with by resorting to a previous situation common conceptual features. In this article, question context case-based (CBR). CBR, plays central role new solved, retrieving cases adapting their solutions. success any CBR system contingent on retrieval case be successfully reused solve problem. We show it often unwarranted assume most also appropriate from reuse perspective. argue must augmented deeper, adaptation knowledge about whether easily modified fit implement idea technique, called adaptation-guided (AGR), which provides direct link between needs. This technique uses specially formulated knowledge, which, during retrieval, facilitates computation precise measure case's requirements. closing, assess broader implications AGR just one growing number methods seek overcome limitations traditional an effort deliver more sophisticated scalable systems.