Heat stress is a significant health problem in tropical work environments such as Indonesia, which can have serious physical and cognitive effects on workers. This study aims to implement the Certainty Factor method in a web-based expert system to address the uncertainty of clinical symptoms in accurately diagnosing heat stress conditions. System knowledge was acquired from general practitioners covering 7 types of diseases and 35 clinical symptoms. The research methods included knowledge acquisition, knowledge modeling, method implementation, and accuracy testing. The evaluation was conducted by comparing the system's diagnosis results with 20 test cases validated by experts. The results showed that the system successfully identified 18 cases correctly and 2 cases incorrectly. Based on these results, this expert system has an accuracy rate of 90%. This accuracy achievement is competitive and in line with previous studies that implemented the Certainty Factor in other health domains. It is concluded that the implementation of this method is feasible and effective in providing an initial diagnosis of heat-related illnesses, thereby assisting healthcare workers and workers in high-risk environments in making appropriate treatment decisions.
Copyrights © 2026