Knowledge Systems Based on Autoepistemic Logic

本研究的目的是调查 自认识逻辑的各种形式的形式化 非单调推理 最近在这方面的工作表明， 非单调的形式主义，如界限，默认逻辑， 作为失败规则的否定可以表示为： 自我认知逻辑 此外，相应的自我认知 形式主义似乎更自然和优雅。 使 自我认知逻辑是一种构造 研究人员将进一步研究基于知识的系统， 其表达能力和寻求专门的有效推理 机制等 这项工作将集中在两个领域，其中 非单调推理经常使用：继承 带有例外和关于动作和时间的推理的层次结构。

The goal of this research is to investigate the applicability of autoepistemic Logic to formalization of various forms of nonmonotonic reasoning. Recent work in this area shows that nonmonotonic formalisms such as circumscription, default logic, and the negation as failure rule can be expressed in autoepistemic logic. Moreover, the corresponding autoepistemic formalisms seem to be more natural and elegent. To make the autoepistemic logic a practical tool for constructing knowledgebased systems the researchers will further investigate its expressive power and seek specialized efficient inference mechanisms. The work will focus on two domains in which nonmonotonic reasoning is frequently used: inheritance hierarchies with exceptions and reasoning about actions and time.