摘要
一、引言基于事例的推理(CBR—Case—Based Reasoning)是通过访问知识库中过去同类问题的求解从而获得当前问题解决的一种推理模式。CBR研究兴起的主要原因是传统的知识处理系统对其“边界”以外的知识处理十分低效,不能从知识的积累中进行学习而达到自我完善,也不能对问题求解的策略与结果提供有效的解释,因而其整体性能十分脆弱。
The paper, firstly, gives a brief introduction to Case-Based Reasoning (CBR). After discussing the goal of the CBR retrieval, the paper summarizes the features and problems of the current CBR retrieval strategies. Based on constraint object knowlege model, a retrieval model, which provides homogenous class hiorarchy for both cases and backgrotmd knowledges, is presented, The paper concludes with the analysis of the characteristics of our retrieval model.
出处
《计算机科学》
CSCD
北大核心
1993年第4期32-35,18,共5页
Computer Science
