| Nowadays multimodal conversational interfaces enables users to communicate with computer systems using a wide range of input/output modalities, such as speech, text, images, gestures etc. Therefore there is a growing need to find not only reliable standardized evaluation methods for such interfaces but also to determine which factors have the highest impact on their quality assessment. The goal of project is to carry out a structural analysis of different multimodal interface layers (speech, text, image) in order to reveal these high impact factors, their relationships and their importance rank. Three case studies are planned with three different conversational systems: a multimodal question-answering interface for medical queries, a multimodal dialogue system for assisting crisis managers and an interactive speech robot for domestic help. Eventually the factors will be incorporated into a taxonomy of quality dimensions for multimodal conversational interactions to provide a common ground on which comparable standardized evaluations can be in the future performed. |
