A estrutura lógico-hierárquica de linguagens de indexação utilizadas por bibliotecas universitárias

Abstract

The use of indexing languages mediates accurately between document and user vocabularies during information retrieval. Nevertheless, it should be updated accordingly to the generation and innovation fo new knowledge. Updating terms depends on the hierarchical structure that guides the inclusion and classification of terms. With the research proposal about hierarchical structures of indexing languages, the bottom-up committee method porposed in the ANSI/NISO-2005 Guidelines for the compilation of controlled vocabularies using the integrated methodological model was adopted, comparing the hierarchies of the Library of Congress Subject Headings (LCSH), the Subject Terminology of National Library (TBN), the BIBLIODATA Language and the USP Vocabulary in the fields of Physics and Mathematics. The results show that BIBLIODATA language does not have certain hierarchical relationships, making any comparison impossible. TBN has eight levels hierarchy for Mathematics and the other languages have four levels for Physics. The conclusion is that it is possible to develop hierarchies using these methods.