Page 111 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42

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

level). “totalConceptAmount” reflects the scale of concept schemes. “topConceptAmount” reflects
the distribution of hierarchical trees in a concept scheme. “level” reflects the depth of knowledge
systems to some extent. “altLableAmount” and “totalConceptAmount” reflect equivalence ratio
indirectly, which is one of the major indicators used to evaluate application performance of thesauri.
The statistic parameters can be extended according to the requirements of thesaurus management.

4 Description of thesaurus microstructure

Thesaurus microstructure is a framework for knowledge description. In order to conform to
semantic web, OWL (Web Ontology Language) was often adopted in early years to describe
thesaurus microstructure so as to transform thesaurus format into ontology format. To describe
knowledge systems more briefly under the framework of semantic web, W3C proposed SKOS
(Simple Knowledge Organization System) in 2009 (World Wide Web Consortium, 2009). SKOS
is a framework of modeling and describing knowledge organization systems based on semantic
web standards. A number of related research studies have been conducted in China (Liu, Zhang,
Pu, Ren, &Mi, 2009; Duan, 2011; Zeng, 2012). In order to transform traditional thesauri to web
thesauri easily, NTW has adopted a partially extended scheme based on SKOS+XL, which is used
to describe thesaurus microstructure.

4.1 Concepts and terms in thesauri

In traditional thesauri, terms are description objects and term relations are used to reveal
associations between concepts. In natural language, there is a phenomenon that one term may
have multiple meanings (ambiguous), and multiple terms may express one meaning (synonym).
In order to resolve the problem of synonym, one of the synonyms can be chosen as preferred term
in traditional thesauri, while others are used as non-preferred terms led to the preferred term via
equivalence relations. To distinguish ambiguous terms, in traditional thesauri, qualifiers are added
to terms to further clarify terms’ meaning. Terms and their qualifiers form “new” terms, so that
term ambiguity can be resolved. Through above approaches, thesauri can perform and achieve their
term control function.
A term is a complex combination of meaning and form. A form may have multiple meanings; a
meaning may have multiple forms. In traditional thesauri, terms are considered as basic component
units of knowledge systems. Thus, if the entanglement of meanings and forms is not effectively
resolve, it will inevitably result in a disorder of semantic relations. In this case, the forms of
particular terms must be changed. Although it helps clarifying term meanings by adding qualifiers,
the artificial terms have lost natural associations with document terms and user language terms.
SKOS data model separates terms’ meanings and forms with meanings as concepts and forms
as literal expressions, namely, labels of concepts. Hierarchical relations and association relations