Estudo de vocabulário controlado na indexação automática: aplicação no processo de indexação do Sistema de Indización Semiautomática (SISA)

• Ano de defesa: 2011
• Autor(a) principal: Narukawa, Cristina Miyuki [UNESP]
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Ciência da informação; Indexação automatica; Sistemas de recuperação da informação; Sistema de indización semiautomática (SISA); Linguagem de indexação; Indexação - Software; Vocabulário controlado; Automatic indexing; Controlled vocabulary; Automatic Indexing Systems; Sistema de Indización Semiautomático (SISA)
• Link de acesso: http://hdl.handle.net/11449/93677

Automatic indexing is a complex process, and delegating the attribution of terms to automatic systems requires analyzing not only the methods, but also the features of indexing instruments. Thereby, we propose to investigate the role of controlled vocabulary in such process, based on the analysis of results from the application of ThesAgro vocabulary in the Semi-Automatic Indexing System (SISA - Sistema de Indización SemiAutomatica -), with the purposes of identifying the characteristics which define and distinguish the types of vocabularies; analyzing methodological proposals and indexing systems; applying the ThesAgro in the SISA, making a comparative analysis related to the manual indexing by the National Library of Agriculture (BINAGRI - Biblioteca Nacional de Agricultura), and analyzing the intervening factors pointing to the occurrence of problems concerning automatic indexing. As a general matter, we seek to contribute to the development of this theme by raising subsidies for adapting controlled vocabularies. We have performed a theoretical review on automatic indexing systems, and an experiment applying the ThesAgro in the SISA, with 100 articles on agriculture , specifically about fruit production. The manual indexing performed by BINAGRI and the comparative analysis with the results from a previous research, which evaluated the performance of the vocabulary from the Health Sciences Descriptors (DeCS - Descritores em Ciências da Saúde) in the before mentioned system, have served as the evaluation parameter. The analysis of results allows us to conclude that the vocabulary conditions the results of the automatic indexing process. Thus, it is necessary to understand it, considering the identification methods of the information representative units, application of linguistic treatment, features of the... (Complete abstract click electronic access below)