摘要
Providing multimodal user interfaces in Ambient Assisted Living scenarios is a challenging task due to large variety of modalities and languages that can be used as well as impairments and preferences of end users. Creating an application that can cope with this multitude of presentation possibilities is highly complex. However, by separating the application from the presentation layer and representing the dialog in an abstract form, it is possible to perform adaptations according to the output parameters. In this work, we present the concept for a Resource Server for multimodal and distributed systems which is capable of storing different kinds of resources and associated metadata, and adapting abstract dialogs. We propose the introduction of a presentation identifier as placeholder for a set of concrete resources, a two-stage mapping between identifiers, and a selection algorithm to cope with the problem of multiple matching resources.