Hershey, Pennsylvania
New York, New YorkBeijing, China
Mental Stress Detection Using Bidirectional Encoder Representations From Transformers

Mental Stress Detection Using Bidirectional Encoder Representations From Transformers

A. Vennila (Kongu Engineering College, India), S. Balambigai (Kongu Engineering College, India), A. S. Renugadevi (Kongu Engineering College, India), and J. Charanya (Kongu Engineering College, India)
Copyright: © 2024 |Pages: 10
DOI: 10.4018/978-1-6684-8531-6.ch013
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

It seems as if people start losing control as they become easily upset, frustrated, and overwhelmed, having problems in resting and quieting their mind, and also feeling bad about themselves, lonely, worthless, and depressed, and avoiding others. If they have experienced the above symptoms, then there is a chance that they are suffering from mental stress. They have to take proper care of their mental health. Stress can be taken care of if it is properly handled and for that detection of stress or the mental state is necessary to provide proper care. The first step in stress detection is sentiment analysis of the users' daily conversations. The authors have proposed an NLP model and have trained it to produce a score for the input ranging between 0 and 1 where 0 is the negative end and 1 is the positive end. The trained model can predict the scores with an accuracy of above 92% on Twitter.
Chapter Preview
Top

L. Zhao et al., (2021) In his study a model named Augmented Education (N = 156) is led, which totals multisource social information spreading over on the web and disconnected learning as well as exercises inside and outside the homeroom. Metrics assessing linear and nonlinear behavioural changes, in particular, can be used to acquire a deeper understanding of the characteristics that lead to outstanding or poor performance. of campus lifestyles are assessed; also, characteristics that characterise dynamic Long short-term memory is used to extract changes in temporal lifestyle patterns (LSTM). Next, a classification technique based onmachine learning is being developed to predict academic success. Finally, visible feedback is being developed to help students (particularly at-risk students) improve their relationships with the university and achieve a better accuracy. Experiments show that the Augment ED model may accurately predict students' academic records.

Complete Chapter List

Search this Book:
Reset