Malleable Access Rights to Establish and Enable Scientific Collaboration

Collaborative systems require access control to prevent unauthorised access and change. Access control has a number of issues, including administration and maintenance overheads. In this paper we argue that it is time to reconsider how access controls work, particularly with scientific and data related domains, and to this end we propose a new paradigm based on a user's demographics and behaviour, rather than simply their identity. In essence, it is both who you are and what you do that is important. We introduce Graft, our Generalised Recommendation Architecture that allows us to support a range of different recommendation models, and provide case studies to illustrate the usefulness of our architecture.