Learning Case Adaptation :: Technology Case-Based Reasoning Essays

Learning Case AdaptationComputer modelings of show window-based conclude (cosmic background radiation) generally guide chance modification using a fixed hardening of accommodation rules. A difficult practical problem is how to identify the intimacy required to guide interlingual rendition for particular labor movements. Likewise, an open issue for CBR as a cognitive model is how carapace accommodation acquaintance is learned. We describe a new approach to acquiring case adjustment companionship. In this approach, adaptation problems are initially solved by reasoning from scratch, using abstract rules about structural transformations and general memory look heuristics. Traces of the processing used for successful rule-based adaptation are stored as cases to modify future adaptation to be done by case-based reasoning. When similar adaptation problems are encountered in the future, these adaptation cases provide task- and domain-specific guidance for the case adaptatio n process. We present the tenets of the approach concerning the relationship between memory search and case adaptation, the memory search process, and the storage and reuse of cases representing adaptation episodes. These points are discussed in the context of ongoing research on DIAL, a computer model that learns case adaptation knowledge for case-based disaster response planning. 1 launching The fundamental principle of case-based reasoning (CBR) for problem-solving is that new problems are addressed by retrieving stored records of prior problem-solving episodes and adapting their solutions to fit new situations. In most case-based reasoning systems, the case adaptation process is guided by fixed case adaptation rules. Practical experience developing CBR systems has shown that it is difficult to establish get case adaptation rules (e.g., Allemang, 1993 Leake, 1994). In defining adaptation rules, a linchpin problem is the classic operationality/generality tradeoff that was first observed in research on explanation-based learning (e.g., Segre, 1987) Specific rules are easy to yield and are reliable, but only apply to a narrow put of adaptation problems abstract rules span a broad range of potential drop adaptations but are often hard and expensive to apply because they do not provide task- and domain-specific guidance. In those CBR systems that do perform case adaptation, specific rules are often used, requiring that the developer perform difficult analysis of the task and domain to determine which rules will be needed. In practice, the problems of defining adaptation rules are so acute that many CBR applications simply ask out case adaptation (e.g., Barletta, 1994). This paper presents a new method by which a case-based reasoning system can learn adaptation knowledge from experience.