Regresar

Peer Feedback Sentiment Analysis Prototype

Abstract:

Higher education institutions currently face several challenges in the face of computerized assessment systems to reach the immersed knowledge of unstructured texts. The application of sentiment analysis through machine learning favors the exploration of unstructured texts for educational management. This article aimed to model a sentiment prediction prototype of the textual feedback generated in the peer assessment process. Was applied the modeling method and tested at the Technical University of Manabi (Ecuador). The results show that it is viable to apply the prototype since most of the students were able to evaluate the task by each rubric criterion, justifying the reasons for placing such numerical scores according to the theme. It is postulated as a helpful tool to assess cognitive, reflective, critical, and collaborative skills in the teaching-learning process.