缺省知识与近似推理

Default Knowledge and Approximate Reasoning

基于模糊逻辑的常识表示与推理是人工智能基础研究的一个新领域，它处理常识的不完全性、不确定性，以及例外情形。本文提出一种新的方法，把缺省逻辑的基本思想推广到近似推理理论中，并用缺省ＡＲ－规则来表示缺省知识。

Commonsense representation and manipulation based on fuzzy logic is a new research field whichhandles the incompleteness, uncertainty and exceptions associated with commonsense knowledge. This paperaims to suggest a way to generalize the basic idea of default logic to the theory of approximate reasoning,inwhich commonsense knowledge is represented in the so called default AR-rules.