ODI Fridays: With great personal data comes great anonymity

Companies across the world think they’re effectively anonymising data about people, but what if someone could reidentify who these people are? Discover the risks and learn about the ODI’s plans to help organisations mitigate those risks. Anonymisation is a set of tools, algorithms and best practices to remove personal information from a dataset while maintaining as much of the data’s utility as possible. In this lecture, Fionntán O’Donnell, the ODI’s Senior Data Technologist, will talk about how to take your data through the anonymisation process in this important, wide-ranging, yet nuanced, subject. In the ODI we believe data should be as open as possible. However, we also believe people who handle that data should do so in a responsible manner. Stewards of data must balance the good of opening and sharing data while handling the risk in sensitive data about people being exposed. About the speaker Fionntán O’Donnell is a Senior Data Technologist and researches how best to use artificial intelligence (AI) and data in fair, accountable and transparent ways. He’s been knocking about researching AI for over a decade now. Initially at the National University of Ireland Galway looking at the semantic web. Then he headed off to Ghent University to study deep learning back before it was even called deep learning. Then to Dublin in the ADAPT Centre, developing machine-learning prototypes in the innovation team. After that he got the boat to London where he worked for the BBC, designing machine-learning systems for News Labs and investigating what AI should mean for a public service broadcaster. He’s called a data technologist because his work is in the intersection between software engineering, research, design, and communication. It can be an odd intersection at times but he’s happy to be there.