A study of usability-aware network trace anonymization

Abstract

The publication and sharing of network trace data is a critical to the advancement of collaborative research among various entities, both in government, private sector, and academia. However, due to the sensitive and confidential nature of the data involved, entities have to employ various anonymization techniques to meet legal requirements in compliance with confidentiality policies. Nevertheless, the very composition of network trace data makes it a challenge when applying anonymization techniques. On the other hand, basic application of microdata anonymization techniques on network traces is problematic and does not deliver the necessary data usability. Therefore, as a contribution, we point out some of the ongoing challenges in the network trace anonymization. We then suggest usability-aware anonymization heuristics by employing microdata privacy techniques while giving consideration to usability of the anonymized data. Our preliminary results show that with trade-offs, it might be possible to generate anonymized network traces with enhanced usability, on a case-by-case basis using micro-data anonymization techniques.

Privacy in the Cloud: A Survey of Existing Solutions and Research Challenges

“…Moreover, these results show that enforcing anonymization procedures following standards lead to a strong reduction of record re-identification. In a study about a hybrid anonymization heuristic [64], Mivule and Anderson aimed towards higher data utility. In the study, the timestamp and Internet Protocol (IP) address were anonymized using distinct methods, and the results demonstrated effective protection against reidentification.…”

“…In a usability awareness study [64], precision degradation is defined as a generalization technique that removes the most precise components of a specific field (e.g., replacing it by 0). With a timestamp, it is possible to have different precision levels (e.g., days, hours, minutes, seconds, and milliseconds).…”

“…The black marker and truncation techniques are special cases of partitioning. A study by Mivule and Anderson's [64] defines permutation as a one-to-one mapping of values. This method is useful when it is necessary to preserve the count or the order of the data sets without maintaining its value.…”

“…Mivule and Anderson [64] claim that random time shifts can be considered a particular case of permutation. When applied to time stamps, it adds a random offset (e.g., seconds or milliseconds) to every record within the timestamp attribute.…”

“…Classification Error Metrics [127] are similar to descriptive statistics as they measure the classification error of an anonymized dataset and compare it to the original data. Mivule and Anderson [64] show that the difference between the two indicators presents a trade-off between data privacy and usability. In this case, both original and anonymized data are passed through a machine learning algorithm that returns a classification error for original and anonymized data.…”