Two approaches to handling proactivity in pervasive systems

A key objective of a pervasive system is to reduce the user's administrative overheads and assist the user by acting proactively on his/her behalf. The aim of this paper is to present some aspects of how proactivity is handled in the approaches used in two different pervasive systems. The Daidalos system provides proactive behaviour based on the assumption that the user is responsible for requesting services and that proactivity is restricted to selecting and personalising these based on the user's preferences. The Persist system uses an extension of this approach combined with an analysis of user intent. The idea behind the latter is that, if the system knows what the user will do next, it can act on the user's behalf, initiating the actions that the user would normally perform. User intent predictions and those produced by the user preferences are used to determine the final action to be taken. © 2010 Springer-Verlag Berlin Heidelberg.