Estudos de Análise de Cocitação de Autores: uma abordagem teórico-metodológica para a compreensão de um domínio

• Autor(a) principal: Grácio, Maria Cláudia Cabrini [UNESP]
• Data de Publicação: 2014
• Outros Autores: Oliveira, Ely Francina Tannuri de [UNESP]
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://inseer.ibict.br/ancib/index.php/tpbci/article/view/137; http://hdl.handle.net/11449/114829
• topic: indicadores normalizados; análise de cocitação de autores; estudos métricos

This piece of research aims at analyzing the absolute and relative co-citation indicators, especially Salton’s Cosine and comparing the contribution of these indicators to the understanding of a domain, applying them to the universe of "Metric Studies" at the BRAPCI base. It also aims at presenting the co-citation network generated from the absolute frequencies and highlighting the groupings of co-cited authors, depending on the relative values, integrating and explaining the information from the two indexes. The domain analysis, by means of its 11 approaches, including “Bibliometric Studies”, focuses on the science characterization and evaluation, in that it allows us to identify and to analyze the conditions under which the scientific knowledge is constructed and socialized. In these studies, the contribution of citation and co-citation analysis is highlighted. Of the total of 147 articles retrieved at BRAPCI base, the authors cited in at least 11 articles, in a total of 38 researchers, were selected. The 38 x 38 symmetric matrix with the absolute frequencies of co-citation and the matrix with the relative values of Salton’s Cosine were generated. The co-citation network with absolute frequency values were constructed, through Ucinet software. Cluster analysis of data with relative values wer performed, using the SPSS software. Significant differences between the absolute and relative indexes, with some high absolute values of co-citation were observed; when considered in relation to the presence of each author, their significance is decreased. As to the generated network, seven groups were determined, in which only one is established for close themes and comes from co-citations in the original sense of the term. Five groups present closeness in absolute and relative indicators. It can be concluded about the importance of the studies of authors' co-citation analysis, which associate the two indexes, absolute and relative, in order to visualize and understand the underlying structures of a scientific domain.