Representations for Multimodal Generation Workshop

Kris Thorisson, Catherine Pelachaud, Hannes Vilhjalmsson, Stefan Kopp

Our Wiki at ISI hosts new follow-up material to the workshop.

The overall goal of this workshop was to further the state of research on multimodal generation by enabling (and getting) people in the field to work together. To this end, we are working towards:
- Framing the problem of multimodal generation in a way allows us to put it into computational models, and that correspond as closely as possibly to real data.

- Defining planning stages of multimodal generation and to identify the knowledge structures that mediate between these stages.

- Rendering these stages and knowledge structures into a framework that lays down modules and interfaces, enabling people to better work together and to use each other’s work with a minimal amount of custom work.

- Evaluating existing specification languages and to assessing them in the context of our new models.

- Formalizing the structures involoved in the model in XML, possibly based on the existing languages.

External project page