International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 10 | Oct 2024
p-ISSN: 2395-0072
www.irjet.net
A PROLIFIC SYSTEM TO DETECT STRESS USING MACHINE LEARNING Prof. Miruna Joe Amali1, Roshni Sri KA2,Priyadharshini T2, Rithika R2 1Professor and Head of the Dept.ofComputerScienceandEngineering,KLNCollegeofEngineering,Pottapalayam,
Sivagangai, Tamil Nadu, India
2B. E Student, Dept. of Computer Science and Engineering, K L N College of Engineering, Pottapalayam
,Sivagangai, Tamil Nadu, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The proposed model presents a comprehensive system for detecting and assessing stress in employees based on text messages, using machine learning techniques. The system utilizes the Naive Bayes algorithm to classify text inputs as stress-related or non-stress-related, and it calculates a stress percentage using both a machine learning model and a keyword-based approach. The analysis is further refined by asking employees to input their weekly working hours and daily sleep duration. These additional factors, coupled with the text analysis, provide a holistic picture of the employee's stress levels. By incorporating machine learning and personal input data, the system ensures a highly accurate and context-sensitive assessment of stress.
productivity and organizational success. Stress, if left unaddressed, can significantly impact an employee's performance, mental health, and overall job satisfaction. To tackle this issue, a comprehensive stress detection and assessment model has been developed, leveraging machine learning to analyze employee text inputs alongside their working hours and sleep patterns. By integrating these factors, the model offers a highly accurate, personalized approach to identifying and managing stress. The core of this system utilizes the Naïve Bayes algorithm to classify text messages as either stress-related or non-stressrelated. This classification is further refined by calculating a stress percentage through a combination of machine learning and a keyword-based analysis. However, the system goes beyond simple text analysis by incorporating additional key factors—employees’ weekly working hours and daily sleep duration. By considering these objective metrics alongside the subjective text input, the model provides a holistic picture of each employee’s stress level. The system evaluates stress across three levels: low, moderate, and high. When stress is detected through text but is mitigated by balanced working hours and sufficient sleep, the system flags a low stress level, indicating that the employee is managing well. If either working hours exceed the company’s set average or sleep falls below 6 hours, moderate stress is identified, and personalized recommendations are provided. In cases where both excessive working hours and poor sleep patterns are present, the stress level is classified as high, triggering immediate attention and intervention. By combining machine learning with real-time data input, this model ensures a context-sensitive and dynamic approach to stress management, aimed at enhancing employee well-being and promoting a healthier work-life balance.
The system evaluates stress at three different levels: low, moderate, and high. If stress is detected in the text, but the working hours remain within the company's set average and the sleep duration is 6 hours or more, the system classifies the stress level as low. In such cases, employees are regarded as managing stress effectively, and the system offers general suggestions to maintain balance based on their working hours. When stress is identified in the text and either the working hours exceed the average or sleep is insufficient (less than 6 hours), the system flags a moderate stress level. In this case, it provides personalized suggestions, helping the employee manage their workload or improve sleep patterns. Finally, if the text indicates stress and both factors— excessive working hours and insufficient sleep— are present, the stress level is labeled high, requiring immediate attention and more focused intervention. If no stress is detected in the text, the system offers suggestions based on working hours alone to promote well-being and maintain a healthy worklife balance. This stress detection system aims to enhance workplace well-being by providing tailored, real-time recommendations that help employees manage stress effectively.
2.LITERATURE SURVEY [1] Analyzing Perceived Psychological and Social Stress of University Students: A Machine Learning Approach - Ishrak Jahan Ratul, Mirza Muntasir Nishat, Fahim Faisal, Sadia Sultana, Ashik Ahmed, Md Abdullah Al Mamun. The concept of the study revolves around using machine learning techniques to better understand the factors contributing to stress among university students. Traditional methods of stress assessment often rely on self-reported surveys,
KEYWORDS: Machine Learning, Naive Bayes Algorithm, Stress Detection, Text Analysis, Employee Well-being
1.INTRODUCTION In today's fast-paced work environment, employee wellbeing has become a critical factor for both individual
© 2024, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 692