Extracting Field Oversees’ Features in Risk Recognition from Data of Eyes and Utterances

Abstract

In this study, we have developed the video based risk recognition training tool with an eye tracking device and a motion sensor. We applied the tool on the risk recognition training in a construction company and extracted features in risk recognition of expert field overseers from their eyes and utterances during the training. As the results of the examinations, typical risk recognition processes for the experts (Meta-knowledge) and the experts' knowledge used for individual risk recognitions (Domain knowledge) were elicited.