<![CDATA[Mental health has become an increasingly important issue amid the growing pressures of modern life, particularly in the workplace. Job demands, performance targets, and the dynamics of social relationships at work can trigger stress that negatively affects employee productivity and well-being. However, low awareness and social stigma surrounding mental health issues often result in stress going undetected at an early stage. This study aims to identify employee stress levels at a company in Batam City using a data mining approach. Data were collected through the distribution of questionnaires based on the DASS-21 (Depression Anxiety Stress Scales), which measures three main aspects: stress, anxiety, and depression. The data were analyzed using the Python programming language, with stages including preprocessing, transformation of scale values into numerical form, and the construction of a classification model using the C4.5 algorithm (Decision Tree Classifier). The results showed that the classification model was able to identify stress levels with an accuracy of 67%. The best performance was observed in the moderate stress class (class 1), with a precision value of 0.71 and a recall of 0.79. In contrast, the classification performance for minority classes such as no stress (class 0) and severe stress (class 2) was relatively low. These findings suggest that the C4.5 algorithm is reasonably effective in recognizing dominant stress patterns but requires further data processing and class-balancing techniques to improve overall model performance. This study is expected to serve as a foundation for early detection and more accurate handling of workplace stress]]>
Copyrights © 2025
