Argumentative paragraphs identification with markers categories
The argumentation in theses is necessary to develop the ideas and convince the reader of the presented statements. However, among university students, there is a deficiency in their ability to argue, which is attended by teachers through the review process. In this article, we present a model for argumentative paragraph detection in sections such as the problem statement, justification, and conclusions. The classification task is performed using machine learning algorithms using five categories of argumentative markers. We present the sets of word patterns for each of the following categories: justification, explanation, deduction, refutation and conditional. Experiments were carried out to evaluate the argumentative paragraph identification reaching encouraging results.