# Scholarly Article Citations in Wikipedia: DOI, PubMed and PMC Identifiers Cited in English Wikipedia (Wikimedia Research)

## Facts

- Provider: Wikimedia Foundation Research (Halfaker, Mansurov, Redi, Taraborelli)
- License: CC0-1.0 (https://figshare.com/articles/dataset/Citations_with_identifiers_in_Wikipedia/1299540)
- Size: 6 files
- Format: Parquet, TSV (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/wikimedia-community/scholarly-article-citations-in-wikipedia
- Access: public, readable without an account
- DOI: https://doi.org/10.6084/m9.figshare.1299540
- Category: Open Data
- Source: https://github.com/halfak/Extract-scholarly-article-citations-from-Wikipedia
- Page: https://scigantic.com/archive/c6e48bb3-8b65-541b-8728-b5c98f58f86b
- Keywords: Open Data, Parquet, TSV, Hugging Face

## About

Every citation to a scholarly article, identified by DOI, PubMed ID (pmid) or PubMed Central ID (pmcid), found in English Wikipedia as of January 2015. That is about 1.84 million (Wikipedia article, scholarly article) pairs. Each row gives the Wikipedia page id and title, the revision id and timestamp at which the citation was first added, the identifier type, and the identifier. Spot checks by the authors found that about 98% of the identifiers resolve. Altmetrics and science-communication researchers use it to measure how research reaches the public encyclopedia and how quickly new papers get cited there. The repository holds the raw TSV (doi-and-pubmed-citations-enwiki-20150112.tsv) and a Parquet copy. A larger, multilingual version is on figshare. The data was released by Aaron Halfaker, Bahodir Mansurov, Miriam Redi and Dario Taraborelli (Wikimedia Foundation) under CC0. Cite Halfaker et al. 2015, Citations with identifiers in Wikipedia, DOI 10.6084/m9.figshare.1299540.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("wikimedia-community/scholarly-article-citations-in-wikipedia")
print(ds)
```

## Data structure

File types in sampled files: Parquet, TSV

Example file: `data/train-00000-of-00001.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/c6e48bb3-8b65-541b-8728-b5c98f58f86b
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/c6e48bb3-8b65-541b-8728-b5c98f58f86b
- Over MCP: `get_schema_card` with id `c6e48bb3-8b65-541b-8728-b5c98f58f86b` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [TSV datasets](https://scigantic.com/topics/tsv)
