Research on human factor accident based on link prediction

Ma Jun and Wan Jie

Abstract

The research on man-machine relationships is an important problem in field of man-machine system. A single factor is considered in much humane error analysis, but very valuable potential factors are not considered in many humane error analysis. In this paper, the link prediction technique is introduced to resolve the multiple factors relation problems associated with the degree of relevance and importance factors index system. The algorithm used to calculate the correlation degree and importance of factors is designed. Finally, the man-machine system of train is used as example of humane error analysis to verify a positive role during of discovering the potential danger by new index system.