Trust: From Cognition to Conceptual Models and Design

Trust as a design issue for information systems has appeared in e-commerce, e-science, and a wide variety of collaborative applications. Much discussion has centred around trust in computational artefacts such as protocols, encryption and security mechanisms; however, little research has focused on exactly what trust means in human terms. In this presentation I will review the psychology literature on trust as a product of reasoning processes, and describe a cognitive model to explain and predict inter-personal and inter-organisational trust. I argue that sound design should be based on cognitive models of users, and these should inform the semantics of conceptual modelling as well as guiding the design process. I will explore the implications of the cognitive model of trust for conceptual modelling in requirements specification languages such as i*. The final part will be more speculative. After a brief review of the implementations of trust-enhancing mechanisms in collaborative and e-science systems, focusing on user interface features rather than encryption, etc. middleware, I will discuss the design challenges for future trustworthy systems. This will cover how trust can be communicated, and issues of honesty when users may not always have the best intentions.