Trustmuse: A model-driven approach for trust management

Abstract

With the increasing acceptance of Trust Management as a building block of distributed applications, the issue of providing its benefits to real world applications becomes more and more relevant. There are multiple Trust Management frameworks ready to be applied; however, they are either unknown to developers or cannot sufficiently be adapted to applications’ use cases. In our research, we have defined a meta model to modularize Trust Management, where each element in the model has clearly defined dependencies and responsibilities – also enforced by a complete API. Based on this model, we were able to develop a process supported by a number of tools that enables non-security expert users to find an applicable Trust Management solution for their specific problem case. Our solution – collectively called the TrustMUSE system – has evolved over an iterative user-centered development process: starting with multiple focus group workshops to identify requirements, and having multiple prototypes to conduct usage observations. Our user evaluation has shown that our system is understandable for system designers, and is able to support them in their work.