Social network data are generally published in the form of social graphs which are being used for extensive scientific research. We have noticed that even a k-degree anonymization of social graph can’t ensure protection against identity disclosure. In this paper, we have discussed how closeness centrality measure can be used to identify a social entity in the presence of kdegree anonymization. We have proposed a new model called k-degree closeness anonymization by adopting a mixed strategy of k-degree anonymity, degree centrality and closeness centrality. The model has two phases, namely, construction and validation. The construction phase transforms a graph with given sequence to a graph with anonymous sequence in such a manner that the closeness centrality measure is distributed among the nodes in a smooth way. The nodes with the same degree centrality are assigned with a closer set of closeness centrality values, making re-identification difficult. Validation phase validates our model by generating
-neighborhood graphs. Algorithms have been developed both for the construction and validation phases.