Efficient generation of ambient intelligent user interfaces

Model-based development is an established approach for generating user interfaces for ambient intelligent environments. Existing approaches neither provide guidelines for creating models nor do they include mechanisms for adapting the user interface to changing user needs during runtime. We present a new approach for tackling these problems by means of integrating HCI patterns into our model-based framework and explicitly capturing changing user characteristics during runtime. The modeling approach presented supports the specification, selection, and instantiation of patterns and the automated elicitation of the users' abilities via interaction analysis in order to reduce usage errors.