120 likes | 263 Views
Architecting a Cross-Disciplinary Thesaurus for the Semantic Web. W. Davenport Robertson National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA Jane Greenberg School of Information and Library Science, University of North Carolina, Chapel Hill, NC, USA
E N D
Architecting a Cross-Disciplinary Thesaurus for the Semantic Web • W. Davenport Robertson • National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA • Jane Greenberg • School of Information and Library Science, University of North Carolina, Chapel Hill, NC, USA • International Conference on Dublin Core and Metadata Applications (DC-2004), Shanghai, China, 14 October 2004
The National Institute of Environmental Health Sciences The Department of Health and Human Services The National Institutes of Health
选择有针对性的最新、最前沿的环境健康新闻、政策及研究综述选择有针对性的最新、最前沿的环境健康新闻、政策及研究综述 以季刊形式发行,发行量37,000份 向读者提供中文简体和繁体两种版本 提供印刷本或上网浏览两种阅读形式 http://ehp.niehs.nih.gov/cehp
Previous Findings on Metadata • Need for subject metadata • Need for authority list of terms • Authors want guidance for subject metadata • Authors would like cataloger assistance • Searchers dissatisfied with surrogates
Needs Assessment for an Environmental Health Thesaurus • “Environmental health” is cross-disciplinary • Environmental health information is crucial for our well-being. • A controlled vocabulary is necessary for retrieving accurate information • No thesaurus adequately covers environmental health today • Recommendation: combine parts of existing thesauri
Thesaurus A, Server 1 Thesaurus D, Server 4 Metathesaurus Ontology Server (Local) Thesaurus B, Server 2 Thesaurus E, Server 5 Thesaurus C, Server 3 Thesaurus F, Server 6 Figure 1. Decentralized Cross-Disciplinary Metathesaurus
Decentralized Semantics Approach for a Metathesaurus • Applied examples • Metadata author • Web site searcher • W3C interoperability principles for shared ontologies (OWL Web Ontology Language) • Advantages of inter-disciplinary model • Integral to the Semantic Web • Distributed resources reduce load on individual components • Individual components can maintain their independence • Components can be re-used in other metathesauri
Decentralized Semantics Requirements • Semantic Web thesaurus standards under development (http://www.w3.org/2004/03/thes-tf/mission) • Open Biological Ontologies example of a consolidated registry (http://obo.sourceforge.net/) • SWOOP example of software that integrates multiple ontologies for searching (http://www.mindswap.org/)
Decentralized Metathesaurus Requirements • Components must be open, preferably open source • Component ontologies must be listed in an ontology registry • Components must use Semantic Web standards • Uniform Resource Identifiers (URIs) • RDF, RDFS, XML • OWL Web Ontology Language
Conclusion • Standards and tools are still under development • Appropriate domain components for an environmental health metathesaurus have to be selected (EPA, NCI, MeSH, etc.) • Organizations have to be convinced of value • Thesauri have to be converted to Semantic Web ontologies • This model can be applied to other cross-disciplinary fields