ABSTRACT
The massive amount of information currently available on the Internet
makes it difficult for teachers to curate quality educational
content or to select material for self-regulated study by students.
Aiming to facilitate these steps in the teaching and/or learning
process this article presents an approach to assist the discovery of
educational content from the hybrid recommendation system and
later classification from the feedback evaluation with sentiment
analysis techniques trained with the Re-Li corpus. This paper describes
the proposed model, the implementation of a prototype and
its application in non-formal training involving 13 participants.