AI2 Reasoning Challenge (ARC) 2018

Summary

• Introduction
• Dataset Structure
• Reference
• License
• Citation

Introduction

A new dataset of 7,787 genuine grade-school level, multiple-choice science questions, assembled to encourage research in advanced question-answering. The dataset is partitioned into a Challenge Set and an Easy Set, where the former contains only questions answered incorrectly by both a retrieval-based algorithm and a word co-occurrence algorithm. We are also including a corpus of over 14 million science sentences relevant to the task, and an implementation of three neural baseline models for this dataset. We pose ARC as a challenge to the community.

Dataset Structure

Data Instances

ARC-Challenge

• Size of downloaded dataset files: 680.84 MB
• Size of the generated dataset: 0.83 MB
• Total amount of disk used: 681.67 MB

An example of 'train' looks as follows.

{
    "answerKey": "B",
    "choices": {
        "label": ["A", "B", "C", "D"],
        "text": ["Shady areas increased.", "Food sources increased.", "Oxygen levels increased.", "Available water increased."]
    },
    "id": "Mercury_SC_405487",
    "question": "One year, the oak trees in a park began producing more acorns than usual. The next year, the population of chipmunks in the park also increased. Which best explains why there were more chipmunks the next year?"
}

ARC-Easy

• Size of downloaded dataset files: 680.84 MB
• Size of the generated dataset: 1.45 MB
• Total amount of disk used: 682.29 MB

An example of 'train' looks as follows.

{
    "answerKey": "B",
    "choices": {
        "label": ["A", "B", "C", "D"],
        "text": ["Shady areas increased.", "Food sources increased.", "Oxygen levels increased.", "Available water increased."]
    },
    "id": "Mercury_SC_405487",
    "question": "One year, the oak trees in a park began producing more acorns than usual. The next year, the population of chipmunks in the park also increased. Which best explains why there were more chipmunks the next year?"
}

Data Fields

The data fields are the same among all splits.

ARC-Challenge

• id: a string feature.
• question: a string feature.
• choices: a dictionary feature containing:
• text: a string feature.
• label: a string feature.
• answerKey: a string feature.

ARC-Easy

• id: a string feature.
• question: a string feature.
• choices: a dictionary feature containing:
• text: a string feature.
• label: a string feature.
• answerKey: a string feature.

Data Splits

name	train	validation	test
ARC-Challenge	1119	299	1172
ARC-Easy	2251	570	2376

Reference

We would like to acknowledge Peter Clark and Isaac Cowhey et al. for creating and maintaining the AI2 Reasoning Challenge dataset as a valuable resource for the computer vision and machine learning research community. For more information about the AI2 Reasoning Challenge dataset and its creator, please visit the AI2 Reasoning Challenge website.

License

The dataset has been released under the Creative Commons Attribution-ShareAlike 4.0 International License.

Citation

@article{allenai:arc,
      author    = {Peter Clark  and Isaac Cowhey and Oren Etzioni and Tushar Khot and
                    Ashish Sabharwal and Carissa Schoenick and Oyvind Tafjord},
      title     = {Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge},
      journal   = {arXiv:1803.05457v1},
      year      = {2018},
}