Using Complex Networks and Natural Language Processing to Characterize and Predict Academic Success

• Ano de defesa: 2025
• Autor(a) principal: Brito, Ana Caroline Medeiros
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Academic success; Cienciometria; Complex networks; Natural language processing; Processamento de linguagem natural; Redes complexas; Scientometrics; Sucesso acadêmico
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/55/55134/tde-24102025-112509/

Scientometrics is an area of study focused on studying Science itself, being common studies in the subareas of citation analysis, scientific mapping, collaboration network analysis, and bibliometrics. During this doctorate, the works developed, although distinct from each other, have as a common objective analyses that enrich our understanding of academic success using mainly complex network techniques and natural language processing as a basis. The works in this collection consisted of: (1) a study on interdisciplinarity, based on citations and references, of researchers and how this measure relates to other measures of visibility and performance; (2) a work about collaborations between authors, defining the top-collaborator of a researcher, we extended this analysis to a large set of researchers, grouping them by research area and observing how the performance measures were impacted by the removal of productions developed with the top-collaborators; (3) application of a framework based on complex networks and natural language processing to two case studies of Chemistry journals, where we were able to obtain overviews of the development of this area; (4) a detailed analysis of online viewing data of articles from the journal PLoS ONE, using statistical analysis and the use of clustering techniques, to characterize these curves, reinforcing the importance of other measures alternative to traditional bibliometrics to understand the dynamics of interest to the academic community. In general, with these works, we expanded our understanding of scientific development from different perspectives, mainly related to academic success measured by citations.