Development of a Proactive Intelligent System for Employee Stress and Well-being Analytics using Machine Learning and Natural Language Processing

Abstract

This research aims to develop and evaluate the efficiency of a proactive intelligent system for analyzing employee stress and well-being by integrating Machine Learning (ML) and Natural Language Processing (NLP) techniques. Grounded in the Job Demands-Resources (JD-R) model, the developed system harmonizes quantitative data-such as overtime (OT) hours and leave statistics-with qualitative insights derived from sentiment analysis of weekly assessments. By utilizing the Random Forest model and Explainable AI (XAI) techniques, the system identifies the root causes of stress and presents the results through an anonymized dashboard to support management decision-making while protecting employee privacy.

           The model performance evaluation revealed that the system achieved an accuracy of 0.87, a precision of 0.85, a recall of 0.89, and an F1-score of 0.87. These results indicate the system’s strong capability in accurately classifying and identifying at-risk employees. In particular, the high recall value demonstrates the model’s effectiveness in minimizing the risk of overlooking employees who are genuinely experiencing stress. Furthermore, the research findings indicate that the system accurately identifies key stressors, particularly a significant positive correlation between OT hours and mental health risk levels. It was observed that during periods where average OT reached 11.5 hours, employee risk levels rose to a "high-risk" threshold (4.3 points). Following a two-month pilot implementation with a sample group of office workers, negative indicators-including fatigue and pressure-decreased from "high" to "manageable" levels. Notably, the feeling of "lack of job control" showed the most significant improvement. Furthermore, the user satisfaction evaluation was high (4.42 points), with users commending the system for its accuracy and its role in promoting organizational well-being. This research represents a vital innovation for proactive human resource management, helping to mitigate burnout and foster sustainable mental health in the workplace.