Natural Language Processing (NLP) to identify the resilience of the return to face-to-face classes at a university

Abstract

The use of technology supported by information is an activity that is increasingly necessary to develop various activities in all fields. The entry into classes of new students at the university after having spent the last years of high school receiving virtual classes causes concern and possible behavioral changes, such is the case of the resilience that can exist when changing from a virtual school environment to a face-to-face university. The objective of this research was to develop a data model that allows sentiment analysis to be carried out with neural networks through Natural Language Processing (NLP), to identify the resilience of the return to face-to-face classes of virtual students at a university, the methodology used was the use of neural networks using natural language processing, through the RISC-10 resilience questionnaire in two modalities, through a Likert scale and through an open question. The results showed that there are differences between what was marked through the questionnaire and what was expressed through the same questions. It is concluded that there is a high difference between what was surveyed and what was described by the students, finding a high resilience when entering classes at the university in person, after developing virtual classes in recent years.