metadata
license: cc-by-sa-4.0
We introduce the Million Authors Corpus (MAC), a novel dataset encompassing contributions from dozens of languages on Wikipedia. It includes only long and contiguous textual chunks taken from Wikipedia edits and links those texts to their authors.
This repository includes the data used for training and testing only. The full MAC dataset is shared in Zenodo repository: https://zenodo.org/records/15538126
Here are few examples: