Abstract
In terms of tourists? attention levels on their travel risk, this research uses the skill of self-organization map(SOM) of artificial neural network to classify tourists and explain their behaviors. Besides discussing the difference of risk items that tourists care when they fell into different attention levels, this study also attempted to explore the relationship between personal characteristics and perceived risk levels. The result revealed that the tourists? perceived risk levels can be recognized with a few of demographic representatives of personal characteristics and showed that the relationship between personal characteristics and perceived risk levels works as a beneficial index of offering customized information service.
