To enhance the management of civil aviation flight safety, a set of factors influencing unsafe behavior is determined through literature analysis and stakeholder survey． Then a network model of relationship among these factors is built by applying social network analysis. Its node attributes, core-edge analysis, structural feature, influence and hierarchical clustering are investigated．The result shows that the key unsafe behavior influencing factor nodes can form a reaction chain, which can in turn connect with others into a high-risk behavior network. Among them, the key unsafe behavior influence factors include job burnout, safety supervision, and business skills proficiency. There is a chain of behavioral reactions, including "the level of knowledge is low- procedures are not strict- civil aviation flight is unsafe", etc. Some kinds of behavior such as job burnout, stress, etc. can serve as bridge nodes in the network of factors influencing unsafe behavior．

civil aviation flight  /  safety management  /  unsafe behavior influence factor  /  social network analysis(SNA)  /  hierarchical clustering