非周期TBox框架下的推理——拟经典描述逻辑与超协调表演算(英文)

Quasi-Classical Description Logics and Paraconsistent Tableau Calculus for Reasoning with Acyclic TBox

源自当今互联网的语义网研究的目的是定义信息语义和网络服务,因此需要赋予网络智能以便能够自动对网络内容进行推理。各种描述逻辑(DLs)在语义网的研究中扮演着重要角色,构成了W3C推荐的网络本体语言(OWL)的基础,而OWL源于为达到语义网目标的人工智能(AI)本体论研究。语义网的知识和数据量巨大、分散、来源众多且因此通常不具有协调性。因此,必须开发针对非协调本体的实用的推理技术。该文基于Hunter的拟经典逻辑(QCL),构造了新型超协调拟经典描述逻辑(QCDLs),避免了逻辑爆炸问题,同时,针对基于非周期TBox的知识库推理问题,建立了QCDLs语义表演算,进而定义了一种可靠、完备且可判定的推理关系,从而构建了完整的语义网推理框架。与其他重要的超协调描述逻辑进行了比较,结果表明QCDLs具有更强的表达语义和推理能力,并且其相关行为在目标层次上表现出经典性。

The forthcoming semantic Web evolving from the current World Wide Web is designed to define the semantics of information and services on the web, thereby endowing the web with intelligence to automatically reason about the web contents. Description logics （DLs） play a substantial role in the semantic Web, since they underlie the W3C-recommended Web ontology language （OWL）, which is derived from ontology research in artificial intelligence （AI） in order to achieve the goal of the semantic Web. However, the knowledge and data in the Semantic Web are large-scale, dispersive, muRi-authored, and therefore usually inconsistent. It is reasonable and imperative to develop practical reasoning techniques for inconsistent ontologies. This paper proposes a new type of paraconsistent description logics based on Hunter＇s quasi-classical logic （QCL）, which are termed as quasi-classical description logics （QCDLs）. QCDLs avoid logical explosion. A semantic tableau calculus is constructed in QCDLs for the reasoning on the knowledge bases with acyclic TBox. Furthermore, a sound, complete and decidable consequence relation based on the calculus is defined. These enable a complete framework for paraconsistent reasoning in the Semantic Web. A comparison with other key paraconsistent description logics is also given. It is shown that QCDLs possess more expressive its connectives behave classically at the object level.