Stress Detection Using Bernoulli Naive Bayes Machine Learning Technique

Stress Detection Using Bernoulli Naive Bayes Machine Learning Technique.

B.Akhil Bharadwaz1
Department of Electronics and Communication Engineering,
Presidency University,
Bangalore, India.
akhilbharadwazbrahmadevi@gmail.com

B.Sasidhar2
Department of Electronics and Communication Engineering,
Presidency University,
Bangalore, India.
bonganisasidhar08@gmail.com

Ms. Sapna R3
Assistant professor, Department of Computer Science and  Engineering,
Presidency University,
Bangalore, India.
sapna.aradhya@gmail.com

Abstract—The lives of people are increasingly being threatened by psychological issues. Before stress develops into a serious health problem, it is critical to recognize and manage it. Social Networking has become a necessary component of our daily lives. On social media sites, people frequently communicate their feelings and emotions, including stress. Social media makes it simple to identify and reduce stressors for both individuals and organizations. Based on user social behavior, stress may be easily identified in users. Also, traditional methods for stress detection are expensive and time-consuming. This study suggests a machine learning method for detecting stress in social media posts using the Bernoulli Naive Bayes algorithm. On a dataset of 2,839 social media posts that includes posts from WhatsApp, Facebook, and Twitter, the proposed approach is assessed, and it yields promising results. We have developed a graphical user interface for the system using Streamlit, which is an open-source app framework that helps us quickly deploy the development into an interactive GUI. This GUI provides the option for the user to enter text. The model determines whether the user is stressed or not based on the input from the user.

Keywords—Bernoulli Naive Bayes Algorithm, CountVectorizer, Stemming, Stress detector, Streamlit.