Datasets:
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- question-answering
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
pretty_name: AirQA
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# AirQA: A Comprehensive QA Dataset for AI Research with Instance-Level Evaluation
|
| 11 |
+
|
| 12 |
+
This repository contains the `metadata`, `processed_data` and `papers` for the **AirQA** dataset introduced in our paper [**AirQA: A Comprehensive QA Dataset for AI Research with Instance-Level Evaluation**](https://www.arxiv.org/abs/2509.16952). Detailed instructions for using the dataset will soon be publicly available in [our official repository](https://github.com/OpenDFM/AirQA).
|
| 13 |
+
|
| 14 |
+
**AirQA** is a human-annotated multi-modal multitask **A**rtificial **I**ntelligence **R**esearch **Q**uestion **A**nswering dataset, which encompasses 1,246 examples and 13,956 papers, aiming at evaluating an agent’s research capabilities in realistic scenarios. It is the first dataset that encompasses multiple question types, also the first to bring function-based evaluation into QA domain, enabling convenient and systematic assessment of research capabilities.
|
| 15 |
+
|
| 16 |
+
## Folder Structure
|
| 17 |
+
|
| 18 |
+
```txt
|
| 19 |
+
metadata/
|
| 20 |
+
|── a0008a3c-743d-5589-bea2-0f4aad710e50.json
|
| 21 |
+
└── ... # more metadata dicts
|
| 22 |
+
papers/
|
| 23 |
+
|── acl2023/
|
| 24 |
+
| |── 001ab93b-7665-5d56-a28e-eac95d2a9d7e.pdf
|
| 25 |
+
| └── ... # more .pdf published in ACL 2023
|
| 26 |
+
└── ... # other sub-folders of paper collections
|
| 27 |
+
processed_data/
|
| 28 |
+
|── a0008a3c-743d-5589-bea2-0f4aad710e50.json # cached data for PDF parsing
|
| 29 |
+
└── ... # more cached data for PDFs
|
| 30 |
+
```
|
| 31 |
+
|
| 32 |
+
Due to Hugging Face's limit on the number of files in a single folder, we packaged `metadata` and `processed_data` into archives.
|
| 33 |
+
|
| 34 |
+
## Dataset Statistics
|
| 35 |
+
|
| 36 |
+
Our dataset encompasses papers from 34 volumes, spanning 7 conferences over 16 years. The detailed distribution is summarized below.
|
| 37 |
+
|
| 38 |
+
<details><summary>👇🏻 Click to view the paper distribution of dataset</summary>
|
| 39 |
+
|
| 40 |
+
| Folder | Conference | Year | Collected |
|
| 41 |
+
| :----: | :--------: | :--: | :-------: |
|
| 42 |
+
| iclr2024 | ICLR | 2024 | 3301 |
|
| 43 |
+
| iclr2023 | ICLR | 2023 | 31 |
|
| 44 |
+
| iclr2020 | ICLR | 2020 | 1 |
|
| 45 |
+
| neurips2024 | NeurIPS | 2024 | 6857 |
|
| 46 |
+
| neurips2023 | NeurIPS | 2023 | 73 |
|
| 47 |
+
| nips2006 | NeurIPS | 2006 | 1 |
|
| 48 |
+
| acl2024 | ACL | 2024 | 161 |
|
| 49 |
+
| acl2023 | ACL | 2023 | 3083 |
|
| 50 |
+
| acl2019 | ACL | 2019 | 1 |
|
| 51 |
+
| acl2019 | ACL | 2016 | 1 |
|
| 52 |
+
| emnlp2024 | EMNLP | 2024 | 55 |
|
| 53 |
+
| emnlp2023 | EMNLP | 2023 | 52 |
|
| 54 |
+
| emnlp2021 | EMNLP | 2021 | 2 |
|
| 55 |
+
| emnlp2013 | EMNLP | 2013 | 1 |
|
| 56 |
+
| icassp2024 | ICASSP | 2024 | 18 |
|
| 57 |
+
| icassp2023 | ICASSP | 2023 | 12 |
|
| 58 |
+
| eacl2024 | EACL | 2024 | 1 |
|
| 59 |
+
| ijcnlp2023 | IJCNLP | 2023 | 1 |
|
| 60 |
+
| arxiv2025 | arXiv | 2025 | 12 |
|
| 61 |
+
| arxiv2024 | arXiv | 2024 | 53 |
|
| 62 |
+
| arxiv2023 | arXiv | 2023 | 61 |
|
| 63 |
+
| arxiv2022 | arXiv | 2022 | 61 |
|
| 64 |
+
| arxiv2021 | arXiv | 2021 | 43 |
|
| 65 |
+
| arxiv2020 | arXiv | 2020 | 25 |
|
| 66 |
+
| arxiv2019 | arXiv | 2019 | 20 |
|
| 67 |
+
| arxiv2018 | arXiv | 2018 | 11 |
|
| 68 |
+
| arxiv2017 | arXiv | 2017 | 6 |
|
| 69 |
+
| arxiv2016 | arXiv | 2016 | 4 |
|
| 70 |
+
| arxiv2015 | arXiv | 2015 | 1 |
|
| 71 |
+
| arxiv2014 | arXiv | 2014 | 1 |
|
| 72 |
+
| arxiv2013 | arXiv | 2013 | 1 |
|
| 73 |
+
| arxiv2012 | arXiv | 2012 | 1 |
|
| 74 |
+
| arxiv2011 | arXiv | 2011 | 1 |
|
| 75 |
+
| uncategorized | - | - | 3 |
|
| 76 |
+
| Total | - | - | 13956 |
|
| 77 |
+
|
| 78 |
+
</details>
|
| 79 |
+
|
| 80 |
+
## Citation
|
| 81 |
+
|
| 82 |
+
If you find this dataset useful, please cite our work:
|
| 83 |
+
```txt
|
| 84 |
+
@misc{huang2025airqacomprehensiveqadataset,
|
| 85 |
+
title={AirQA: A Comprehensive QA Dataset for AI Research with Instance-Level Evaluation},
|
| 86 |
+
author={Tiancheng Huang and Ruisheng Cao and Yuxin Zhang and Zhangyi Kang and Zijian Wang and Chenrun Wang and Yijie Luo and Hang Zheng and Lirong Qian and Lu Chen and Kai Yu},
|
| 87 |
+
year={2025},
|
| 88 |
+
eprint={2509.16952},
|
| 89 |
+
archivePrefix={arXiv},
|
| 90 |
+
primaryClass={cs.CL},
|
| 91 |
+
url={https://arxiv.org/abs/2509.16952},
|
| 92 |
+
}
|
| 93 |
+
```
|