论文部分内容阅读
针对一阶逻辑在复杂结构数据环境中存在模式搜索空间庞大和不能发明新谓词的缺点,提出了使用类型化的高阶逻辑知识表示语言Escher去表示各种复杂结构的数据,利用其强类型语法有效地约束知识发现过程中模式的搜索空间和高阶的特点去解决新谓词构造的问题。设计了以Escher为基础的复杂结构数据中的知识发现过程和基于复杂结构数据的聚类算法,并以实验验证了其有效性。
In order to solve the shortcomings of first-order logic in the complex structured data environment, such as large search space and no new predicate, this paper proposes to use the typed higher-order logical knowledge representation language Escher to represent data of various complicated structures. Using its strongly typed syntax Effectively constrains the pattern search space and high-order features in the process of knowledge discovery to solve the problem of new predicate construction. The knowledge discovery process based on Escher-based complex structure data and the clustering algorithm based on complex structure data are designed, and its effectiveness is verified by experiments.