Análisis de sentimientos en los mensajes recibidos en el entorno virtual de aprendizaje de la modalidad abierta y a distancia de la UTPL

Abstract

Abstract: This research work has as an objective to analyze and classify the text messages that students of the Open and Distance Mode of UTPL sent to the EVA during the academic period April/2019-August 2019. The primary purpose of this work is to know the level of positivism or the negativism of the students. Different text mining methods are identified and applied to calculate the polarity of the messages. For example: with R using the Tidy library; in java the stanford.nlp framework; with Python the methods were executed; Naive Bayes and support vector machine, the Textblob library was implemented as well. The process has several phases that start after the collection of information from the EVA. Then, the messages are stored in a repository to perform the data cleaning, in which all those characters that do not contribute to the analysis are eliminated. Finally, the "Sentiment Analysis" is executed applying authomatic learning methods. In the consolidated result, it was observed that there were 84.14% positive messages, 12.32% negative and finally 3.54% of neutral messages giving a high level of positivism.