In connection with security of databases or social networks in order to anonymize them the background knowledge of intruders is of enough importance. These knowledges need to be framed properly in order to make them a part of the anonymization process.

What are the different approaches followed in this connection? Any reference to source materials will be useful.