The rapid development of Web2.0 technology promotes the online social network which is becoming the most popular platform for people to spread information. The huge amount of data released by users brings huge commercial value and privacy information disclosure. To solve the problem that the privacy information flow cannot be controlled in online social network, an information flow control model based on neighbor node intimacy is proposed. By computing the sensitivity of resources that users allow their friends to access, the model measures the intimacy relationship of neighbor nodes. And the number of common neighbors between users and their friends can be used to improve the model. In addition, information is divided into different intimacy security levels by referring to the ideas of the level of multilevel security (MLS). By setting reasonable scope of intimacy to convey information, social network managers may control the privacy information flow within a certain scope. The simulation experiments with parameter adjusting demonstrate the validity and practicability of the proposed model.