It’s significantly meaningful for improving information transmission in social network site(SNS) to detect the most influential nodes. The key nodes who occupy structural holes are easier to acquire information benefits and control benefits from their relationship network. This study summarizes the existing structural holes measuring methods based on Burt’s structural holes theory, and provides a new improved method according to the features of SNS, which considers the degree of nodes and the topological structure of their neighbor nodes. In addition, to evaluate the accuracy of those measuring methods, the susceptible-infected-recovered(SIR) model is used to simulate experiment in four real data sets of social network. The experimental result shows that our method has a better accuracy and applicability. This paper comes up with a new idea and method for detecting key nodes who occupy structural holes in SNS.