# Deal or No Deal? End-to-End Learning for Negotiation Dialogues

## TL;DR

This paper introduces a large dataset and end-to-end models for negotiation dialogues, demonstrating that AI can learn complex communication and reasoning skills in semi-cooperative settings without annotated states.

## Contribution

It presents the first end-to-end training approach for negotiation dialogues, including a novel dialogue rollout technique to improve planning and performance.

## Key findings

- End-to-end models can learn negotiation skills from raw dialogue data.
- Dialogue rollouts significantly enhance model performance.
- Public dataset and code are available for further research.

## Abstract

Much of human dialogue occurs in semi-cooperative settings, where agents with different goals attempt to agree on common decisions. Negotiations require complex communication and reasoning skills, but success is easy to measure, making this an interesting task for AI. We gather a large dataset of human-human negotiations on a multi-issue bargaining task, where agents who cannot observe each other's reward functions must reach an agreement (or a deal) via natural language dialogue. For the first time, we show it is possible to train end-to-end models for negotiation, which must learn both linguistic and reasoning skills with no annotated dialogue states. We also introduce dialogue rollouts, in which the model plans ahead by simulating possible complete continuations of the conversation, and find that this technique dramatically improves performance. Our code and dataset are publicly available (https://github.com/facebookresearch/end-to-end-negotiator).