Mineração de textos e gestão do conhecimento: aplicação na experiência operacional em geração de energia nuclear nas Usinas de Angra I e II

Abstract

The knowledge is important differential for the organizations to achieve its aims. It to have two dimensions: tacit and explicit. The documents, reports, releases, articles and free texts are accessible forms to most of persons for become explicit the yours topics of knowledge. In last two decades, especially after intensive use of digital mídia and Internet technologies, the organizations noticed the quickly and great growth of digital text databases. In the organizations that to use great level of knowledge in the yours operations and management, the text documents to have great value while objects of information and knowledge. However, to extract and to use this knowledge is great match, especially in large databases of documents. The knowledge discovery in text databases text mining is one powerful resource to explore the universe of knowledge enable in digital textual collections and can be to build strategic views of reality elements to set of texts. This is achieving with computation resources and approaches of Information Science for support Knowledge Management in organizations, especially those use of intensive knowledge. To produce relevant information and to generate new and strategic knowledge is the great aim of text mining. This thesis studies these questions, to set analysis of organization reality ELETRONUCLEAR, enterprise of power energy of nuclear fonts, that to have large database of digital textual document to concern of activities of generate nuclear energy. Difbrain®, is text mining software that has been use in process text analysis of ELETRONUCLEAR text databases, to doing intelligent reading of documents and extracting relationships, patterns, relevant contents in texts, across semantic analysis, use of Information Retrieval and Artificial Intelligence technologies in synoptic studies about of set of documents.