Network data is of increasing interest for official statisticians. Publishing such data introduces re-identification risks that differ from tabular data or microdata records, since the network structure may yield revealing clues about the identity of a node.
We show that our definition has some desirable properties and we evaluate this measure on a small scale-free network. We point out some avenues for improving upon the current computational complexity of the algorithms implementing the anonymity measure, and discuss the most important avenues for future research.
