Friday lunchtime lecture: Data federations – collaborate without sharing

Data assets have more value when used collaboratively than on their own. For many organisations, however, this kind of collaboration is currently very difficult, whether because the data they handle is private or otherwise sensitive, or they lack the resources and know-how to embark on large-scale cooperative projects. The concept of ‘data federations’ offers a model which allows multiple parties to commit their digital assets to a shared project whilst retaining control over the extent to which their data is exposed, as well as reducing the need for standardisation. This is achieved through a flexible, purpose-oriented approach to collaboration building combined with the deployment of privacy-preserving technologies such as distributed data mining or federated learning, which allow insights or predictive models to be generated from data without requiring it to be gathered in a central location. In this talk, writer and researcher Richard Woodall will describe how, by offering an alternative path towards collaboration, a ‘data federation’ opens the benefits to a new range of people, allowing smaller and less digitally mature organisations the chance to develop shared value from their digital assets, cultivate cooperative relationships, and build their capacity for strategic coordination. About the speaker Richard Woodall is a writer and researcher with an interest in the culture and political economy of digital technologies. For the past year his work at Etic Lab has focussed on the theory and practice of building Data Federations