Social network analysis is an important methodology in sociological research. Although social network data is very useful to researchers and policy makers, releasing it to the public may cause an invasion of privacy. In this paper, we generalize the techniques used to protect private information in tabulated data, and propose some safety criteria for assessing the risk of breaching confidentiality by releasing social network data. We assume a situation of data linking, where data is released to a particular user who has some knowledge about individual nodes of a social network. We adopt description logic as the underlying knowledge representation formalism and consider the safety criteria in both open-world and closed-world contexts.