Análise de dados de saúde: mineração de texto com a utilização do Orange Canvas para exploração da informação

Abstract

The present work analyzes the textual data contained in birth plan documents through a text mining tool. The study was developed to find patterns in the reports of pregnant women contained in birth plan documents in the field of obstetric health. Orange Canvas was used to perform text mining techniques. The objective was to use the information contained in a Birth Plan database in the free text field to extract knowledge about the specific characteristics of the obstetric history of each pregnant woman, in relation to the experiences reported by the pregnant woman, as positive and negative experiences. The interpretation of the data was performed considering the look of the researcher with that of the professional specialist of the Hospital das Clinicas of the Federal University of Minas Gerais, in order to validate the applied model. Organizations have been concerned with getting to know their data better while information science tends to study them to better support clinical decision making. There are significant gains in applying the text mining tool to unstructured data, and it is an efficient way to analyze and exploit the data.