---
task_categories:
- text-classification
- text-generation
language:
- en
---

# Dataset Card for ArXivEdits (Sentence-aligned documents)

<!-- Provide a quick summary of the dataset. -->

ArXivEdits is a dataset comprising 751 English scientific papers from arXiv, each with sentence alignments across multiple revisions. 
It also includes fine-grained, span-level edits which are annotated with the revision type and the underlying intention for 1000 sentences.
This dataset consists of *only the sentence-aligned documents subset* of the whole dataset. The annotated edits can be found in [this dataset](https://huggingface.co/datasets/miwytt/arXivEdits_edits).


### Dataset Sources

<!-- Provide the basic links for the dataset. -->

Check the original paper and repository for a more detailed description:

- **Repository:** https://github.com/chaojiang06/arXivEdits/tree/main
- **Paper:** https://aclanthology.org/2022.emnlp-main.641/

- **Licenses:**
  - http://arxiv.org/licenses/assumed-1991-2003/
  - http://arxiv.org/licenses/nonexclusive-distrib/1.0/
  - http://creativecommons.org/licenses/by-nc-sa/4.0/
  - http://creativecommons.org/licenses/by-sa/4.0/
  - http://creativecommons.org/licenses/by/3.0/
  - http://creativecommons.org/licenses/by/4.0/
  - http://creativecommons.org/licenses/publicdomain/
  - http://creativecommons.org/publicdomain/zero/1.0/


## Citation

<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->

```bibtex
@article{jiang-etal-2022-arXivEdits,
  title={arXivEdits: Understanding the Human Revision Process in Scientific Writing},
  author={Jiang, Chao and Xu, Wei and Stevens, Samuel},
  journal={In Proceedings of EMNLP 2022},
  year={2022}
}
```