Decision-Theoretic Question Generation for Situated Reference Resolution: An Empirical Study and Computational Model

TL;DR

This paper investigates how dialogue agents can effectively ask clarification questions to resolve referential ambiguity in situated environments, supported by empirical analysis and a new computational decision-theoretic model.

Contribution

It introduces a novel decision network model for clarification questions that outperforms baseline methods in ambiguous environments, based on empirical dialogue data.

Findings

Distribution of question types used for ambiguity resolution

Influence of dialogue-level factors on reference resolution

Model outperforms slot-filling baseline in varying ambiguity environments

Abstract

Dialogue agents that interact with humans in situated environments need to manage referential ambiguity across multiple modalities and ask for help as needed. However, it is not clear what kinds of questions such agents should ask nor how the answers to such questions can be used to resolve ambiguity. To address this, we analyzed dialogue data from an interactive study in which participants controlled a virtual robot tasked with organizing a set of tools while engaging in dialogue with a live, remote experimenter. We discovered a number of novel results, including the distribution of question types used to resolve ambiguity and the influence of dialogue-level factors on the reference resolution process. Based on these empirical findings we: (1) developed a computational model for clarification requests using a decision network with an entropy-based utility assignment method that operates across modalities, (2) evaluated the model, showing that it outperforms a slot-filling baseline in environments of varying ambiguity, and (3) interpreted the results to offer insight into the ways that agents can ask questions to facilitate situated reference resolution.