Affiliations: [a] College of Information Science and Engineering, Northeastern University, Shenyang, China | [b] School of Software, Northeastern University, Shenyang, China
Correspondence:
[*]
Corresponding author: Prof. Z.M. Ma, College of Information Science and Engineering, Northeastern University, Shenyang, 110819, China, Tel./Fax: +86 24 83681582; E-mail: [email protected].
Abstract: The Semantic Web aims at creating ontology-based and machine-processable Web content, and thus the success and proliferation of the Semantic Web depends on quickly and cheaply constructing Web ontologies. Currently, huge amounts of data are stored in database systems. Therefore, constructing ontologies by extracting domain knowledge from databases will facilitate the development of Web ontologies. In this paper, we develop a formal approach and an automated tool for constructing ontologies from object-oriented database models (OODMs). Firstly, we propose a kind of formal definition of OODMs and give the semantic interpretation of OODMs, where we consider both the structural and dynamic aspects of OODMs. Then, we give the ontology definition, in which the ontology is formulated in OWL DL language and consists of the ontology structure and instances. On this basis, we propose a formal approach that can transform an OODM and its corresponding database instances into the ontology structure and instances, respectively. Furthermore, following the proposed approach, we implement a prototype transformation tool called OODB2OWL, which can automatically construct ontologies by extracting knowledge stored in object-oriented databases db4o. Finally, based on the constructed ontologies, and in order to show the perspectives on how the constructed ontologies may be useful for improving some database applications, we investigate how to reason on OODMs (e.g., consistency, satisfiability, subsumption, and redundancy) by the reasoning mechanism of ontologies, which would help to improve some database applications and construct ontologies more exactly.
