2008 GraphBasedRemAndReasforOntologies

From GM-RKB
Jump to: navigation, search

Subject Headings: Graph-based Knowledge Representation, Conceptual Graph Theory, Ontology.


Notes

Cited By

Quotes

Abstract

An ontology, in the Knowledge Engineering and Artificial Intelligence sense, is a framework for the domain knowledge of an intelligent system. An ontology structures the knowledge, and acts as a container for the knowledge. We define knowledge conjunction as one or more agents using multiple ontologies to perform tasks and understand the domain. Once a common ontology is agreed upon, the agents then have a common background in which to share knowledge. No current method exists that allows intelligent agents to agree on a common framework for sharing knowledge, although there has been some work in comparing semantic meanings within an ontology [44]. This means that agents are unable to use the knowledge of another agent, as the knowledge is meaningless if it isn’t presented in a proper context or a common ‘language’.

In this Chapter, we first give an overview of Conceptual Graph Theory, including what conceptual graphs are and how they work. We then take a different point-of-view for the representation of ontologies. Rather than constructing a CG to represent the ontology, we assert that the CG formalism is better exploited by using a combination of the concept type hierarchy, the canonical formation rules, the conformity relation and subsumption to act as the framework for the knowledge base. An unpopulated ontology (which is simply a framework for the knowledge) is represented by the type hierarchy without specific individuals, while the populated ontology (the framework, as well as the knowledge of the domain) is represented by a hierarchy and the specific conceptual graphs which instantiate individuals, constraints, situations or concepts.,


 AuthorvolumeDate ValuetitletypejournaltitleUrldoinoteyear
2008 GraphBasedRemAndReasforOntologiesDan R. CorbettGraph-based Representation and Reasoning for OntologiesStudies in Computational Intelligencehttp://dx.doi.org/10.1007/978-3-540-78293-310.1007/978-3-540-78293-32008