The approach presented in this paper is based on the field of human resource management with the aim to extend the analysis of human resources by a graph theory perspective with an output representation by synthetic social networks. Further analysis of human resources is focused on their division into communities with similar competencies and skills. We used betweenness concept of centrality for finding important persons in the network that share their skills and competencies with workers in other communities and can therefore serve as contact persons between communities with different skills. This method can also be used for suggesting worker team composition based on similarity of workers’ skills for different roles.