Abstract: Quantitative behavior pattern analysis of workers and its application to operation are very significant for team management. In this paper, clustering analysis of workers and management rules are implemented respectively via social network analysis method. The results indicate that the divergence of teams' cohesion is significatively high, and the behavior of some individual workers has great influence on the cohesion of the team as a whole. To management system, the rules in the same category have similar performance, but some solo rules have disparate performance to others in the same category. Comparing with the management system, the behavior of works has smaller minimum variance grouping coefficient, which indicates that it can be grouped more equably. Above quantitative evaluation can be used in supervising the behavior of workers and improving the management rules.