KNOWLEDGE EXTRACTION FROM SCIENTIFIC ARTICLES: A PROPOSAL FOR INITIAL RESEARCH METHODOLOGY ON TEXT MINING

• DOI: 10.22533/at.ed.317272210053

• Palavras-chave: Bibliometrics; Web of Science Base; Automatic extraction; Vosviewer software.

• Keywords: Bibliometrics; Web of Science Base; Automatic extraction; Vosviewer software.

• Abstract:

  The need to extract textual information encourages the development of automated tools for document reading, pattern recognition and knowledge extraction. The large amount of scientific content available on the internet makes it difficult to search for the most relevant and up-to-date information. In this work, a research methodology is performed to support the process of information retrieval in a scientific environment with the initial theme of Text Mining. Steps were defined for searching, visualizing and filtering data, through computational tools, with the objective of finding relevant scientific material with assertiveness. The method resulted in a list of 49 published articles with an average impact factor of 2.193, material that was compiled and used for initial mapping of the topic. It is concluded that the initial research methodology proposal provides results with a reduced time, with relevance to the material and that the use of general-purpose computational tools, even with a lot of manual intervention, helps to define the initial state of a scientific research. Text mining can make the extraction of knowledge from scientific research already produced more productive, promoting the identification of the state of the art and research gaps.

• Número de páginas: 13