The Wikimedia Foundation has partnered with Kaggle to launch a new dataset optimized for artificial intelligence model training, specifically to combat the scraping of Wikipedia content by AI developers. This dataset, available in English and French, includes various elements from Wikipedia articles while being openly licensed. By providing this well-structured JSON data, Wikimedia aims to create an appealing alternative to scraping Wikipedia, which has been taxing its infrastructure due to high bot traffic. The collaboration aims to enhance accessibility for smaller entities in the AI and data science communities.
Wikimedia partners with Kaggle to release a machine-learning optimized dataset to dissuade AI developers from scraping Wikipedia for data.
Collection
[
|
...
]