Swedish Dialogue Systems
Our aim in this project is to bring together research groups in Sweden who have competence in speech analysis and synthesis, dialogue structure and the interpretation of natural language and move them towards the creation of a common environment which will allow for the creation of generic dialogue systems and the investigation of fundamental issues in dialogue technology. For this to be possible a number of problems, ranging from speech recognition to interaction design, have to be solved. In this project we focus on the higher system levels: dialogue management, robust utterance interpretation, including use of prosodic information, and effective utterance planning. In order to ensure flexibility and ecological validity of dialogue systems we will also create a platform for multimodal spoken language corpora which will contain spoken language audio and video recordings with standardized transcriptions. We believe that the results of a project of this type will be very valuable both as a resource for future research and as a way of raising Swedish competence in the area of multimodal spoken language processing.