2021 年 5 巻 2 号 p. 31-40
Prediction of complaint causes is important to prevent complaints for good service. STAMP/STPA is widely used for prediction of accident causes in hardware-based safety critical system domains. This research proposes an enhanced STAMP/STPA for prediction of complaint causes in the service domain, named Service STAMP/STPA. We expand control action of the original STAMP/STPA to social action focusing on human factors. Typical guide words are also added. Finally we compare the proposed method to the original one in a manned parking space guidance service. Our method predicted a greater number of complaint causes with higher severity. We conclude that the Service STAMP/STPA is effective for complaint prediction in the service domains.