Page 101 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42

100 Journal of Library Science in China, Vol. 8, 2016

(2) OCLC vocabulary mapping project
In this project, nine specific mapping tasks were carried out using DDC or LCSH as the mapping
targets, and six more vocabularies were added, including Education Resources Information Center
(ERIC) vocabulary, Library of Congress Classification (LCC), and Medical Subject Headings
(MeSH), and so on (Vizine-Goetz, Hickey, Houghton, & Thompson, 2004; Si, Xu, Wu, & Chen,
2007). The reasons for using DDC/LCSH as the mapping centers are similar to HILT project.
(3) KoMoHe project
KoMoHe project (Mayr & Petras, 2008) has integrated 25 vocabularies commonly used in social
sciences. In this project, a vocabulary is mapped to other vocabularies if necessary. As a result,
three vocabularies became the actual centers in the mapping web.
(4) UMLS
Unified Medical Language System (UMLS) has integrated 178 vocabularies. This system
consists of three knowledge sources: metathesaurus, semantic network, specialist lexicon and
lexical tools (U.S. National Library of Medicine, 2015; D. Y. Li, et al, 2012). Metathesaurus is
a concept layer above the specialist lexicon which contains a variety of member vocabularies.
Terms from the member vocabularies are integrated into concepts and then into metathesaurus.
Above the metathesaurus there is a semantic network used as a category system for concepts in
metathesaurus. Therefore, UMLS has already formulated its own concepts, not only mapping
relations as presented and discussed in the previous projects.
（5）Scientific and Technical Knowledge Organization System (STKOS)
As a part of a project sponsored by the 12 Five-Year National Scientific and Technical Support
th
Plan of China, and titled “Building and Application Demonstration of Knowledge Organization
System Oriented English Scientific and Technical Literature”, the STKOS project initiated by the
National Scientific and Technical Library (NTSL), is the largest thesaurus semantic integration
project in China thus far.
STKOS has integrated more than one hundred English thesauri, which covers almost all
scientific disciplines, including natural sciences, engineering and technologies. In terms of
semantic integration mode, STKOS is similar to UMLS. Its logical structure includes four bottom-
up layers, namely, material layer, basic lexicon, concept layer, and category system. The material
layer contains special thesauri, classifications, dictionaries, glossaries, etc. Some thesauri can be
chosen from the material layer and imported into basic lexicon. Synonyms from different thesauri
in the basic lexicon are compared and integrated into one, which then included in the concept
layer. Concepts in the concept layer inherit the original relations in the source thesauri, but the
relations are not further processed. The category system is a classification scheme for concepts in
the concept layer. Concepts can be aggregated further according to their discipline characteristics.
The category system in STKOS is different from the one in UMLS. The semantic network
in UMLS is a formal classification system. Its hierarchical structure strictly follows semantic
constraints for genus-species relationship. Each semantic type has a specific definition. The