用CERLEL方法表达知识

Knowledge Representation by CERLEL Approach

本文简单介绍了ER模型(美国数据库专家PeterP.Cher教授的实体—关系模型),强调了用它描述客观世界,尤其是面向结构对象知识的可行性。接着阐述了知识表示中用我国杰出的逻辑学家林邦谨的制约逻辑(简称LEL)取代正统数理逻辑的必要性和可能性。最后作者提出了一种通用的知识表示模式——CERLEL方法(实体—关系模型和制约逻辑相结合的方法)。提出CERLEL方法的目的是为知识库系统的设计提供一种通用工具。

A general knowledge representation methodology called CERLEL (Combination of Entity-Relationship Model and Lin's Entailment Logic) is presented in this paper. CERLEL Approach is based on LEL. LEL can precisely portray the portion of the real world without any sementic distortion. Especially, it can describe possible events and inevitable events, to do 'possible, necessary' reasoning, inductive inference and analogous inference. The CERLEL Approach absorbs the advantages of ER model, for example, using the diagram to portray the real world can be very deep and very subtle, so it is easy for one to understand the conceptual schema as a whole. Because of these two reasons, CERLEL Approach can describe entities, entity-relations, also can des- cribe various logic relationships between entity relations, so it is a powerful and general methodology to represent knowledge and to do inference. Besides, using CWA (Closed World Assumption) is safe in the system based on CERLEL Approach, which also has been argued.