Utilização de processamento de linguagem natural para identificação do domínio da escrita formal em redações da língua portuguesa

Abstract

In Brazil, the main means of entering a public or private university is through the National High School Exam, ENEM. This exam requires that the candidate has the ability to write a good dissertation-argumentative text according to the formal norm of the Portuguese language, and can be eliminated from the exam if he does not fulfill this requirement. In order to help the candidate to identify his mistakes and help in the process of writing a good essay, this article proposes the implementation of a tool capable of identifying the spelling and grammatical errors of a text using techniques of Natural Language Processing (PLN). The analysis of the tools showed that the results obtained by the research are promising, mainly in relation to the identification of grammatical errors.