Reference-Centric Models for Grounded Collaborative Dialogue

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 2130–2147 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 2130 Reference-Centric Models for Grounded Collaborative Dialogue Daniel Fried† Justin T. Chiu‡ Dan Klein† †Computer Science Division, UC Berkeley ‡Department of Computer Science, Cornell Tech {dfried,klein}@cs.berkeley.edu, jtc257@cornell.edu Abstract We present a grounded neural dialogue model that successfully collaborates with people in a partially-observable reference game. We focus on a setting where two agents each observe an overlapping part of a world context and need to identify and agree on some object they share. Therefore, the agents should pool their information and communicate pragmatically to solve the task. Our dialogue agent accu- rately grounds referents from the partner’s ut- terances using a structured reference resolver, conditions on these referents using a recurrent memory, and uses a pragmatic generation pro- cedure to ensure the partner can resolve the ref- erences the agent produces. We evaluate on the OneCommon spatial grounding dialogue task (Udagawa and Aizawa, 2019), involving a number of dots arranged on a board with con- tinuously varying positions, sizes, and shades. Our agent substantially outperforms the previ- ous state of the art for the task, obtaining a 20% relative improvement in successful task completion in self-play evaluations and a 50% relative improvement in success in human eval- uations. 1 Introduction In grounded dialogue settings involving high de- grees of ambiguity, correctly interpreting and in- formatively generating language can prove chal- lenging. Consider the collaborative dialogue game shown in Figure 1. Each player has a separate, but overlapping, view on an underlying context. They need to communicate to determine and agree on one dot that they share, and both players win if they choose the same dot. To succeed, each participant must—implicitly or explicitly—ground their part- ner’s descriptions to their own context, maintain a history of what’s been described and what their partner is likely to have, and informatively convey parts of their own context. We present a grounded pragmatic dialogue sys- tem which collaborates successfully with people Underlying Hidden State Agent’s View Partner’s View A: I have two large light grey dots with a smaller darker dot below and to the right of it P: I have one large lightest grey dot in my entire view A: I have two large light grey dots P: There is a smaller slightly darker grey dot next to the lightest grey and larger dot A: Yes , let’s pick the light one P: The light grey and large dot? A: Yes it ’s the one . Let’s pick that one P: SELECT red A: SELECT blue Figure 1: An example dialogue produced by our system (A) with a human partner (P). The participants have dif- ferent but overlapping views of a shared board, which contains dots of different shapes and sizes. The part- ners must collaborate through dialogue in order to find and select a dot that is visible to both. on the task above. Figure 1 shows a real example …

Content truncated to 500 words