Visual Attention-Based Access: Granting Access Based on Users' Joint Attention on Shared Workspaces

During collaboration, individual users' capacity to maintain awareness, avoid duplicate work and prevent conflicts depends on the extent to which they are able to monitor the workspace. Existing access control models disregard this contextual information by managing access strictly based on who performs the action. As an alternative approach, we propose managing access by taking the visual attention of collaborators into account. For example, actions that require consensus can be limited to collaborators' joint attention, editing another user's personal document can require her visual supervision and private information can become unavailable when another user is looking. We prototyped visual attention-based access for 3 collaboration scenarios on a large vertical display using head orientation input as a proxy for attention. The prototype was deployed for an exploratory user study, where participants in pairs were tasked to assign visual attention-based access to various actions. The results reveal distinct motivations for their use such as preventing accidents, maintaining individual control and facilitating group awareness. Visual attention-based access has been perceived as more convenient but also less certain when compared to traditional access control. We conclude that visual attention-based access can be a useful addition to groupware to flexibly facilitate awareness and prevent conflicts.