Managing uncertainty in dialogue information state for real time understanding of multi-human meeting dialogue

Introduction

Human speech processing is riddled with ambiguity and uncertainty on a number of levels: e.g. uncertainty of speech-processing; lexical and structural ambiguity in parsing; dialogue–act classification; intention recognition and interpretation. Information-state approaches to dialogue management typically only maintain a single current state and utilize strategies for resolving ambiguities and uncertainty immediately they arise.

We are concerned with tracking and understanding dialogue between multiple human participants—specifically, in meetings—in such a way that the dialogue system does not intervene. In this scenario, the system is not able to provide feedback on whether or not it has understood, and is unable to ask for clarification or ambiguity resolution. Our ultimate aim is to model human-human dialogue (to the extent that it is feasible) in real-time, providing useful services (e.g. relevant document retrieval) and answering queries about the dialogue state and history (e.g. “what action items do we have so far?”). Our approach has been to extend our existing dialogue system, based on the information-state update approach— which supports a rich semantic interpretation of multi-utterance constructions—to cope with the added uncertainty inherent in two-person meetings in which the participants speak, point, and draw on a whiteboard.