Abstract

The need for well-thought-out anonymisation has never been more acute. In the current data environment, increasing numbers of large and linked datasets are created about people. At the same time there are political, social and economic drivers encouraging greater sharing of anonymised data. Add into the mix that anonymisation is a complex process and then it is perhaps not surprising that the drive to share data has led to some ill thought out and poorly-anonymised data publications including the Netflix[1], AOL [2] and New York taxi[3] cases, underlining how important it is to carry out anonymisation properly and what can happen if you do not. In this presentation, I introduce a framework for undertaking well thought out anonymisation: the Anonymisation Decision-making Framework (ADF). The ADF attempts to unify the technical, legal, ethical and policy aspects of anonymisation to provide a comprehensive guide to doing anonymisation in practice.