The Challenges of Effectively Anonymizing Network Data

The uncertainties that currently exist about the efficacy of network data anonymization, from both technical and policy perspectives, leave the research community in a vulnerable position. Even as the field marches forward, it does so with little understanding of the implications of publishing anonymized network data on the privacy of the networks being monitored and the utility to researchers. Without that understanding, data publishers are left to wonder what fields must be anonymized to avoid legal fallout, while researchers question the confidence of results gained from the data. However, the extensive work done on micro- data anonymity provides the network research community with several useful insights about how to effectively apply anonymization to published data. At the same time, prior wisdom cannot be applied directly without first overcoming several challenges, including the development of appropriate privacy and utility definitions for the more complex case of network data. Addressing these challenges is essential, in our view, to ensure the continued, yet responsible, availability of network trace data to support security research.