Ambient Intelligence Interaction via Dialogue Systems

The vision of Ambient Intelligence (AmI), that expresses a paradigm in information technology, is based on the increasing technological advances in embedding computational power, information and sensing capabilities into everyday artefacts and environments (Ducatel et al., 2001). Intelligent environment is a technological concept that, according to Mark Weiser, is "a physical world that is richly and invisibly interwoven with sensors, actuators, displays, and computational elements, embedded seamlessly in the everyday objects of our lives, and connected through a continuous network" (Weiser, 1991). Consequently, the computational model of this kind of environments can be analysed as a large collection of networked artefacts. The use of embedded systems to control artefacts, tools and appliances has been common practice for almost two decades now. With every new generation, these controllers provide an ever increasing list of capabilities in the form of assistance, information, and customization. However, it is the addition of communication capabilities that changes the perspectives of what such systems can do: gather information from other sensors, real objects and computers on the network, or enable user-oriented customization and operations through short-range communication (Cook & Das, 2004). For this, AmI technological infrastructures must be able to spontaneously reconfigure themselves and grow from the available, purposeful artefacts in order to become effective in the real world. Recent approaches are based on exploiting the affordances of real artefacts by augmenting their physical properties with the potential of computer based support (Streitz, 2007). Combining the best of both worlds requires an integration of real and virtual worlds resulting in hybrid worlds. Despite the current availability of technology, there is a notorious absence of large scale settings. In this context, the fundamental question is what kind of intelligent interfaces is needed to access a large federation of artefacts within an AmI scenario. In this scenario, each one of the hybrid artefacts or controllable resources should be designed to allow plug and play integration in an AmI multimodal architecture (Dahl et al., 2008). 5