Proposal for the Clustering of Characteristics to Identify Emotions in the Development of a Foreign Language Exam

Abstract

Automatic emotion identification allows for obtaining information on emotions experienced by an individual during certain activities, which is essential for improving their performance or preparing for similar experiences. This document aims to establish the clusters of variables associated with the identification of emotions when a group of students takes a foreign language exam in Portuguese. Once the data clusters are determined, it is possible to establish the perception of emotions in the students with relevant variables and their respective decision thresholds. This study can later be used to build a model that relates the measured variables and the student’s performance so that strategies can be generated to help the student achieve better results on the test. The results indicate that the clusters and range values of the variables can be obtained to observe changes in the concentration of the students. This preliminary information can be used to design a fuzzy inference system to identify the student’s state of concentration.