Towards Distribution Options in the End-User Development of Multi-device Mashups

In recent years several approaches addressed the development of applications for multi-device environments. Due to the mobility of devices, the volatile availability of their resources, e.g., limited battery life or communication quality, and changing user requirements, the application developer cannot anticipate every situation, e.g., new or leaving devices, their capabilities and interaction resources, at the application's design time. Hence, end-users should be able to create and customize multi-device applications at run time. Available solutions already provide user-centered development tools, but they are not sufficient and lack intelligent assistance to offer possible design options, in particular in terms of the application's distribution. In this paper, we present our ongoing work to provide assistance in the end-user development of multi-device mashups (MDM) by recommending distribution options with respect to available devices, their capabilities and resources to ease the user-driven mashup development. Furthermore, we investigate the use of aggregation rules and composition ranking rules to rate distribution options regarding their quality properties and their context fitness, e.g., when the MDM is used at home or in an automotive environment.