---
task_categories:
- text-retrieval
language:
- en
size_categories:
- 100K<n<1M
---

### Dataset Summary

A dataset of 166K million arXiv articles for applications like trend analysis, paper recommender engines, category prediction, co-citation networks, knowledge graph construction and semantic search interfaces.
We extract a max of 1000 papers from each catgeory to build the dataset. If a category does not have 1000 papers the max amount of papers present are used.

### Data Fields

- `id`: ArXiv ID (can be used to access the paper)
- `authors`: Authors of the paper
- `title`: Title of the paper
- `journal-ref`: Information about the journal the paper was published in
- `doi`: [Digital Object Identifier](https://www.doi.org)
- `abstract`: The abstract of the paper
- `categories`: Categories / tags in the ArXiv system

```
@misc{clement2019arxiv,
    title={On the Use of ArXiv as a Dataset},
    author={Colin B. Clement and Matthew Bierbaum and Kevin P. O'Keeffe and Alexander A. Alemi},
    year={2019},
    eprint={1905.00075},
    archivePrefix={arXiv},
    primaryClass={cs.IR}
}
```