Presenter/Author Information

A. Simón-Cuevas
L. Ceccaroni
A. Rosete-Suárez

Keywords

semantic analysis, knowledge sharing, concept map, ontology, owl

Start Date

1-7-2008 12:00 AM

Abstract

In order to obtain semantic interoperability in open systems, the systemsinvolved need to agree on commonly understandable knowledge representations. The caseconsidered in this paper is one of a system needing to communicate and share knowledgeusing concept maps and ontologies, in the context of the environmental sciences and thesemantic Web. We present how to formally obtain ontologies codified in the OWLlanguage from concept maps. Concept maps are a flexible and informal form of knowledgerepresentation, while OWL is a language oriented to processing carried out by machines.The mapping between concept maps and ontologies is a formal transformation, whichsemantically analyzes the relations linking the concepts in the map. The proposed methodincludes a concept-sense disambiguation procedure and uses the WordNet lexicalknowledge base. It also includes automatic learning of the semantics of the relationsbetween concepts. The proposed method has been applied to the environmental-knowledgedomain, through tests of several concept maps with labels in Spanish.

Share

COinS
 
Jul 1st, 12:00 AM

An approach to formal modeling of environmental knowledge via concept maps and ontologies

In order to obtain semantic interoperability in open systems, the systemsinvolved need to agree on commonly understandable knowledge representations. The caseconsidered in this paper is one of a system needing to communicate and share knowledgeusing concept maps and ontologies, in the context of the environmental sciences and thesemantic Web. We present how to formally obtain ontologies codified in the OWLlanguage from concept maps. Concept maps are a flexible and informal form of knowledgerepresentation, while OWL is a language oriented to processing carried out by machines.The mapping between concept maps and ontologies is a formal transformation, whichsemantically analyzes the relations linking the concepts in the map. The proposed methodincludes a concept-sense disambiguation procedure and uses the WordNet lexicalknowledge base. It also includes automatic learning of the semantics of the relationsbetween concepts. The proposed method has been applied to the environmental-knowledgedomain, through tests of several concept maps with labels in Spanish.