Privacy Metrics

Along with a colleage - Dr. Yoan Miche - we presented a paper outlining ideas regarding using mutual information as a metric for establishing some form of  'legal compliance' for data sets. The work is far from complete and the mathematics is getting horrendous!

The paper entitled "On the Development of A Metric for Quality of Information Content over Anonymised Data-Sets" was presented at the excellent Quatic 2016 conference held in Lisbon, Sept 6-9, 2016.

We were also extremely fortunate in that a presented in our session didn't turn up and we were graciously given the full hour not just to present the paper but give a much fuller background and details of future work and the state of our current results.

Here are the slides:

On The Development of A Metric for Quality of Information Content over Anonymised Data Sets from Ian Oliver

Abstract:

We propose a framework for measuring the impact of data anonymisation and obfuscation in information theoretic and data mining terms. Privacy functions often hamper machine learning but obscuring the classification functions. We propose to
use Mutual Information over non-Euclidean spaces as a means of measuring the distortion induced by privacy function and following the same principle, we also propose to use Machine Learning techniques in order to quantify the impact of said obfuscation in terms of further data mining goals.

Citation:

Ian Oliver and Yoan Miche (2016) On the Development of A Metric for Quality of
Information Content over Anonymised Data-Sets. Quatic 2016, Lisbon, Portugal, Sept 6-9, 2016.

Requirements Engineering and Privacy

A lot of travelling this month to conferences and speaking about privacy engineering (as usual). I just spent a week in Beijing at RE'16 (Requirements Engineering 2016) where I both presented a paper on privacy requirements and participated in a panel session on digitalisation and telecommunications - more on that later.

Anyway, here are the slides from the privacy paper:

Experiences in the Development and Usage of a Privacy Requirements Framework from Ian Oliver

And here is the abstract:

"Any reasonable implementation of privacy requirements can not be made through legal compliance alone. The belief that a software system can be developed without privacy being an integral concept, or that a privacy policy is sufficient as requirements or compliance check is at best dangerous for the users, customers and business involved. While requirements frameworks exist, the specialisation of these into the privacy domain have not been made in such a manner that they unify both the legal and engineering domains. In order to achieve this one must develop ontological structures to aid communication between these domains, provide a commonly acceptable semantics and a framework by which requirements expressed at different levels of abstractness can be linked together and support refinement. An effect of this is to almost completely remove the terms ‘personal data’ and ‘PII’ from common usage and force a deeper understanding of the data and information being processed. Once such a structure is in place - even if just partially or sparsely populated - provides a formal framework by which not only requirements can be obtained, their application (or not) be justified and a proper risk analysis made. This has further advantages in that privacy requirements and their potential implementations can be explored through the software development process and support ideas such as agile methods and ‘DevOps’ rather than being an ‘add-on’ exercise - a privacy impact assessment - poorly executed at inappropriate times."

Ian Oliver (2016) Experiences in the Development and Usage of a Privacy Requirements Framework. Requirements Engineering 2016 (RE'16), Beijing, China, September 12-17, 2016