Dutch News Summaries Translated from CNN and Daily Mail

Encompassing Dutch translations of the CNN/Dailymail news summarization dataset. It was created by the ml6team using a Helsinki-NLP machine translation pipeline and was last updated in October 2022.

Use Cases

• Fine-tune a Dutch text summarization model using the translated article and summary pairs.
• Train a sequence-to-sequence model for machine translation quality assessment between English and Dutch.
• Analyze the linguistic features of machine-translated news text in Dutch.

Strengths

• Derived from the established CNN/Dailymail dataset, a standard benchmark for summarization.
• Uses the Helsinki-NLP/opus-mt-en-nl model, a known translation pipeline.
• Last updated in October 2022.

Limitations

• The dataset is a machine translation, which may introduce errors or unnatural phrasing not present in the original English.
• No information is provided on the number of rows, columns, or sample data for validation.
• The temporal and geographic coverage of the original English articles is unspecified.

Provenance

Source

Translated from the original CNN/Dailymail dataset hosted on Hugging Face.

Collection Method

Created using a Hugging Face translation pipeline (Helsinki-NLP/opus-mt-en-nl model).

Freshness

Last updated 2022-10-22.