As the technical capabilities of latest mobile devices are combined with mobile broadband internet access, we are ready to make use of free and natural speech in mobile services by utilizing optional and complementary means of input. In these kinds of multimodal dialogue systems the possibility to disambiguate between several input and also output modalities can substantially increase intelligibility of dialogues and robustness of interaction. The combination of natural speech and tactile gestures as input mediums, especially in map-based systems, have shown promising results, although mature commercial applications are still to be developed. In this paper we present the MUMS Multimodal Route Navigation System, which allows users of a mobile device to present public transportation route queries with any preferred combination of speech and pen input, with the system providing navigational information via speech and graphical map representations. We also present the results of a user study, with focus on the users' subjective experience of the MUMS system and especially its multimodal qualities.
展开▼