Dataset Card for Deal or No Deal Negotiator

Dataset Summary

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.

Supported Tasks and Leaderboards

Train end-to-end models for negotiation

Languages

The text in the dataset is in English

Dataset Structure

Data Instances

{'dialogue': 'YOU: i love basketball and reading THEM: no . i want the hat and the balls YOU: both balls ? THEM: yeah or 1 ball and 1 book YOU: ok i want the hat and you can have the rest THEM: okay deal ill take the books and the balls you can have only the hat YOU: ok THEM: ', 'input': {'count': [3, 1, 2], 'value': [0, 8, 1]}, 'output': 'item0=0 item1=1 item2=0 item0=3 item1=0 item2=2', 'partner_input': {'count': [3, 1, 2], 'value': [1, 3, 2]}}

Data Fields

dialogue : The dialogue between the agents. input : The input of the firt agent. partner_input : The input of the other agent. count : The count of the three available items. value : The value of the three available items. output : Describes how many of each of the three item typesare assigned to each agent

Data Splits

train	validation	test
dialogues	10095	1087	1052
self_play	8172	NA	NA

Dataset Creation

Curation Rationale

[More Information Needed]

Source Data

Initial Data Collection and Normalization

[More Information Needed]

Who are the source language producers?

[More Information Needed]

Annotations

Annotation process

[More Information Needed]

Who are the annotators?

Human workers using Amazon Mechanical Turk. They were paid $0.15 per dialogue, with a $0.05 bonus for maximal scores. Only workers based in the United States with a 95% approval rating and at least 5000 previous HITs were used.

Personal and Sensitive Information

[More Information Needed]

Considerations for Using the Data

Social Impact of Dataset

[More Information Needed]

Discussion of Biases

[More Information Needed]

Other Known Limitations

[More Information Needed]

Additional Information

Dataset Curators

[More Information Needed]

Licensing Information

The project is licenced under CC-by-NC

Citation Information

@article{lewis2017deal,
  title={Deal or no deal? end-to-end learning for negotiation dialogues},
  author={Lewis, Mike and Yarats, Denis and Dauphin, Yann N and Parikh, Devi and Batra, Dhruv},
  journal={arXiv preprint arXiv:1706.05125},
  year={2017}
}

Contributions

Thanks to @moussaKam for adding this dataset.