Data Science Seminar: Anonymisation, Risk and Personal Data Event
For more information regarding this event, please email Dr Elena Simperl at E.Simperl@soton.ac.uk .
Event details
Anonymisation is the process of removing information from a dataset in order that people represented in it are not identifiable, thereby respecting privacy, confidentiality and data protection. However, a series of high-profile data breaches, where individuals were reidentified from anonymised datasets - AOL search data, Netflix preference data, open data about New York taxi journeys - have amply demonstrated the mathematical truth that anonymisation is only irreversible at the cost of removing all utility from the data. Anonymisation is one way to share personal data legally, but in a world characterised by big data, is it impossible to do it safely? This talk argues that anonymisation needs to be understood, not as a technical fix, but as a process that takes place within a social, legal, technological and economic context. Only when we take this wider and more realistic view can we properly assess the contribution it can make to legal and ethical data science.
Speaker information
Dr Kieron O'Hara,is an associate professor and principal research fellow within the WAIS group, working on the SOCIAM project. His interests are the social and political aspects of computing and the Web, especially in terms of privacy and trust. The influence of his report on open data and privacy for the Cabinet Office in 2011 was reflected in the coalition government's open data white paper of 2012. He chaired an advisory panel for the Ministry of Justice and the Home Office on the safe release of crime and criminal justice data, and is one of the leads of the UKAN network of anonymisation professionals, set up by the Information Commissioner's Office.