摘要
元知识是描述、使用一般知识的知识,具有变化的特点.提出了用变换产生式,即可拓知识作为元知识的一种新的表示形式.它比一般用规则式表示的元知识更适应变化环境.在解决矛盾问题中用新元知识描述解决矛盾的本质是目标变换(TG)或条件变换(TL),引起了关联函数值(TK(P))的变换;知识发现中用新元知识说明属性约简变换Treduc是依赖于算子ASGF(Ci)的计算结果,基于粗糙集的知识获取是依赖于算子Aupdow(Ei,Yj)的计算结果;专家系统中用新元知识作为控制知识推理的运行更体现了变化的特点.这些应用例子表明了用变换产生式作为元知识的一种新的表示形式更适应变化环境.
Meta-knowledge is a form of knowledge for describing and applying general knowledge. It is in a continually changing state. The authors proposed a new method to represent meta-knowledge using transform-production, or extension knowledge. Compared with traditional recta-knowledge representational methods using rule implication, extension knowledge is more suitable for changing environments. In solving contradictory problems with new recta- knowledge, in essence, goal-transform （TG） or condition-transform （TL ） triggers the transform of a related function＇ s value （ TK（P） ）. The application of new recta-knowledge to knowledge discovery in a database shows that the transform of attribute reduction （ Tredue） depends on the computational result of operator ASGF （Ci）. Knowledge acquisition based on a rough set is dependent on the computational value of operator Aupdow（Ei, Yj）. Extension knowledge is also suitable for representing recta-knowledge in order to control the operation of expert systems. It is obvious that the new method, which employs transform-production to represent recta-knowledge, plays an instructive role in summarizing and solving such problems.
出处
《智能系统学报》
2009年第4期331-334,共4页
CAAI Transactions on Intelligent Systems
基金
国家自然科学基金资助项目(70671031)
关键词
元知识
变换产生式
可拓知识
矛盾问题
知识发现
专家系统
meta-knowledge
transform production
extension knowledge
contradictory question
knowledge discovery in database
expert system
