Aplicação de ferramentas de análise de sentimentos em um ambiente de aprendizagem colaborativa - Collabora

Abstract

Sentiment Analysis is an area of study that uses natural language processing to extract, format, and study data according to your feeling, which can improve and improve the quality of various fields in society. Within collaborative and collective learning systems, analytics can improve feedback and outcomes for participating students and faculty. To improve the collaboration of each student within the Collabora system, a virtual collaborative learning object that focuses on improving each student's learning by delivering tools that enable the exchange of ideas between them, in this work the analysis of feelings in the messages that students exchanged within the project. This process was done through research using the database generated by the system itself, messages exchanged between students during activities, and with the aid of tools such as NLTK and spaCy and lexical dictionaries for formatting and analyzing data using the lexical method. With the calculated results, the collaboration calculation is reformulated to use the feeling of each student as a condition of their commitment to the activity, and then carry out a comparison between the tools and the collaboration calculation carried out in Collabora. The work can contribute to real-time integration in learning systems or even create one more parameter for the student's involvement with the activity.