Toward a Model for Incremental Grounding in Dialogue Systems

Spoken dialogue systems have until recently upheld the simplifying assumption that the conversation between the user and the system occurs in a strict turn-by-turn fashion. In order to have more human-like, uent conversations with computers, a new generation of spoken dialogue systems has arisen that is capable of processing the user’s speech in an incremental way. As the user is speaking, the automated speech recognizer produces transcriptions of what the user has been saying so far with small increments, typically every 200ms. Based on these transcriptions, the system can decide to perform overlapping behaviors such as head nods, backchannels and the like. We have studied the AMIMeeting Corpus in order to identify ways of grounding in humanhuman dialogue that a system would be able to pick up using incremental processing. ese incremental grounding behaviors include overlapping feedback, the completion of un nished utterances that were initiated by the other party and responding to an utterance before it is completed. We have developed an incremental groundingmodel that supports those incremental grounding behaviors. e input of the model consists of incremental hypotheses of the explicit and predicted content of the utterance, i.e. what has been uttered so far and what is likely to be the full utterance meaning respectively. We have de ned how grounding acts can be identi ed incrementally and how the grounding state, i.e. the collected contents and progress of the common ground units (CGUs), is updated accordingly. We de ned new types of acknowledgments and how they affect the content of the CGU they ground, e.g. answering an un nished question also grounds the part of the question that was not uttered. We implemented our model in the SASO4 dialogue system as a proof-of-concept of our approach, showcasing an up-to-date grounding state through the execution of a simple overlapping feedback policy.