Generation models for spoken dialogues

The paper discusses what kind of genera-tion model is suitable for spoken dialogue re-sponses. We describe different existing models of generation, and compare them from the point of view of spoken dialogue systems. We intro-duce a model of generation based on new infor-mation focus, and argue that this addresses the communicative requirements of flexible spoken dialogue systems (incrementality, immediacy and interactivity). We discuss the relationship between the dialogue manager and the gener-ator, and what kind of interface they should share. We describe a flexible shallow genera-tion approach which combines template-based generation with a pipeline of distinct process-ing levels.