Document Type

Journal Article

Department/Unit

Department of Computer Science

Title

Dynamic sub-ontology evolution for traditional Chinese medicine web ontology

Language

English

Abstract

As a form of important domain knowledge, large-scale ontologies play a critical role in building a large variety of knowledge-based systems. To overcome the problem of semantic heterogeneity and encode domain knowledge in reusable format, a large-scale and well-defined ontology is also required in the traditional Chinese medicine discipline. We argue that to meet the on-demand and scalability requirement ontology-based systems should go beyond the use of static ontology and be able to self-evolve and specialize for the domain knowledge they possess. In particular, we refer to the context-specific portions from large-scale ontologies like the traditional Chinese medicine ontology as sub-ontologies. Ontology-based systems are able to reuse sub-ontologies in local repository called ontology cache. In order to improve the overall performance of ontology cache, we propose to evolve sub-ontologies in ontology cache to optimize the knowledge structure of sub-ontologies. Moreover, we present the sub-ontology evolution approach based on a genetic algorithm for reusing large-scale ontologies. We evaluate the proposed evolution approach with the traditional Chinese medicine ontology and obtain promising results.

Keywords

Ontology, Traditional Chinese medicine, Sub-ontology, Knowledge, Evolutionary computation

Publication Date

10-2008

Source Publication Title

Journal of Biomedical Informatics

Volume

41

Issue

5

Start Page

790

End Page

805

Publisher

Elsevier

Peer Reviewed

1

Funder

This work is partially supported by subprogram of China 973 Project (No. 2003CB317006), National Science Fund for Distinguished Young Scholars of China NSF Program (No. NSFC60533040), National Key Technology R&D Program (No. 2006BAH02A01), a NSFC program under Grant No. NSFC60503018, and also a grant from ChinaGrid (CNGI-04-15-7A).

DOI

10.1016/j.jbi.2008.05.008

Link to Publisher's Edition

http://dx.doi.org/10.1016/j.jbi.2008.05.008

ISSN (print)

15320464

ISSN (electronic)

15320480

This document is currently not available here.

Share

COinS