DocumentCode :
2918673
Title :
Structure-Based Ontology Evaluation
Author :
Ning, Huang ; Shihan, DIAO
Author_Institution :
Sch. of Comput. Sci. & Eng., Beihang Univ., Beijing
fYear :
2006
fDate :
Oct. 2006
Firstpage :
132
Lastpage :
137
Abstract :
In order to build high quality ontologies, ontology evaluation technologies are needed. Now, most evaluation methods focus on the syntax evaluation which guarantees the correctness and completeness of ontology. The primary goal of these evaluation methods is to prevent applications from using inconsistent or incorrect ontologies. During last few years, some studies pay attention to the common meaningful structures among ontologies, which are analogous to each other. However, fewer studies concentrate on the internal structure of the ontology. Ontology, as the representation of knowledge, should have similar structure with domain knowledge. Furthermore, the ontology with well-organized structure will make it easy to understand, learn, apply, and reuse. Therefore, based on the statistics and the graph theory, this paper puts forward an improved ontology evaluation method, in which six properties are proposed to describe the characteristics of ontology structure. After discussing the algorithm of these six properties, we introduce an ontology development tool, in which the evaluation method has been integrated. This tool supports the entire ontology development process, and can help ontology developers refine the structure of ontology, and easily make a decision that which aspect of the ontology should be revised or improved
Keywords :
graph theory; ontologies (artificial intelligence); statistics; domain knowledge; graph theory; knowledge representation; ontology structure evaluation; statistics; Application software; Buildings; Computer science; Graph theory; Knowledge engineering; Ontologies; Semantic Web; Statistics;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
e-Business Engineering, 2006. ICEBE '06. IEEE International Conference on
Conference_Location :
Shanghai
Print_ISBN :
0-7695-2645-4
Type :
conf
DOI :
10.1109/ICEBE.2006.97
Filename :
4031643
Link To Document :
بازگشت