Dataset origin: https://www.kaggle.com/datasets/dhruvildave/wikibooks-dataset
This is the complete dataset of contents of all the Wikibooks in 12 languages. The content contains books of the following languages: English, French, German, Spanish, Portuguese, Italian and Russian, Japanese, Dutch, Polish, Hungarian, and Hebrew; each in its own directory. Wikibooks are divided into chapters and each chapter has its own webpage. This dataset can be used for tasks like Machine Translation, Text Generation, Text Parsing, and Sematic Understanding of Natural Language. Body contents are provided in both newline delimited textual format as would be visible on the page along with its HTML for better semantic parsing.
Refer to the starter notebook: Starter: Wikibooks dataset
Data as of October 22, 2021.
- Downloads last month
- 34