MURML: A Multimodal Utterance Representation Markup Language for Conversational Agents

Alfred Kranstedt, Stefan Kopp and Ipke Wachsmuth
This paper presents work on an artificial anthropomorphic agent with multimodal interaction abilities. It focuses on the development of a markup language, MURML, that bridges between the planning and the animation tasks in the production of multimodal utterances. This hierarchically structured notation provides flexible means of describing gestures in a form-based way and of explicitly expressing their relations to accompanying speech.
Multimodal Communication, Form-based Gesture Description, Markup Languages, Conversational Agents
PDF (~249 k)
Anke Weinberger, 2003-02-20, 2003-06-03