DEBATE ON ANONYMISATION

Debate on Anonymisation with Yves MontjoyeMark Elliot, Natalie Shlomo, Keith Spicer & Magnus Rattray, Robert McCombe

DEBATE PROPOSITION: “DATA CAN EITHER BE USEFUL OR ANONYMISED BUT NEVER BOTH.”

This session was something of a departure from the usual IDSAI seminar format.

Much of the data that data scientists work with is about, or concerns, people. Therefore, legal compliance and best ethical practice is crucial to ensure that we meet confidentiality assurances privacy impact is minimised. Anonymisation is one approach to dealing with these issues. However, its reliability, value and even its definition are hotly contested.

We therefore arranged this debate between four speakers; each with their own view of what anonymisation is and how useful it is in our increasingly complex data environment. Paul Comerford from the Information Commissioner’s Office summarised the debate before the audience voted on the debate proposition.

  • Moderator/chair: Magnus Ratray(University of Manchester)
  • Speaker 1: Yves de Montjoye (Imperial College London)
  • Speaker 2: Keith Spicer (Office for National Statistics)
  • Speaker 3: Natalie Shlomo (University of Manchester)
  • Speaker 4: Mark Elliot (University of Manchester)
  • Concluding remarks: Robert McCombe (Information Commissioners Office)

Programme

Introduction by the moderator

1: Opening statements (The four speakers) 40 minutes

Each speaker makes a 8-10 minute opening statement

2: Rejoinders 10-20 minutes

Each then replies to the others’ statements and asks questions of others. 2-4 minutes each.

3: Open discussion 45 minutes

The audience enters the debate; speakers go to panel mode.

4: Concluding remarks (Paul Comerford): 5-10 minutes

5: Vote 5 minutes

The Audience will then vote on the proposition.