Skip to content

Dataset · Babelscape

Multilingual corpus for named entity recognition

Babelscape/wikineural

From 1 to 10 million annotated texts in six European languages for the NER task, collected using BabelNet and BERT.

What's inside

A large corpus ranging from 1 to 10 million examples for the token-classification task (named entity recognition, NER). The data covers six languages: English, German, Spanish, French, Italian, and Dutch. The annotation was obtained automatically based on the BabelNet multilingual knowledge base and BERT family transformers.

What you can do with it

  • Training and fine-tuning of multilingual NER models
  • Pre-training or evaluation of token classifiers in European languages
Tasks
Token classification
Languages
German, English, Spanish, French, Italian, Dutch, Polish, Portuguese and 1 more
Size
1M–10M rows
Downloads this month
5.4K
Likes
38

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Babelscape/wikineural")