论文部分内容阅读
ＭＫＬ是知识获取系统ＮＤＫＡＳ中实现的一个元知识学习算法，它在分类及抽象的基础上归纳出二叉树结构的元知识，用以有效地组织知识库中的规则．ＭＫＬ生成的元知识满足元知识的基本性质．本文给出了ＭＫＬ的算法描述，基本性质的满足性证明及算法的应用例子．
MKL is a meta-knowledge learning algorithm implemented in NDKAS, which summarizes the meta-knowledge of binary tree structure based on classification and abstraction to effectively organize the rules in the knowledge base. The meta-knowledge generated by MKL satisfies the basic nature of meta-knowledge. This paper presents the algorithm description of MKL, proof of the basic nature of satisfaction and application of the algorithm.