Turkish News Articles for Abstractive Summarization

Encompassing Turkish news articles for abstractive text summarization, with a size category of 100K to 1M entries. It was created by Batuhan Baykara and Tunga Güngör, with the associated research paper published in 2022.

Use Cases

• Fine-tune a sequence-to-sequence model for abstractive text summarization on Turkish news articles.
• Train a text classification model using the provided 'Task Categories' tags to categorize news content.
• Analyze linguistic patterns specific to the Turkish language ('Language:tr') for natural language processing research.

Strengths

• Dataset size is categorized as 100K to 1M entries, indicating a substantial collection.
• Specifically designed for abstractive summarization, a key task for agglutinative languages like Turkish.
• Associated with a peer-reviewed research paper published in Language Resources and Evaluation in 2022.

Limitations

• Exact row count, column structure, and sample data are unavailable, hindering initial assessment.
• Specific geographic coverage within the 'Region:us' tag is unclear and may not represent all Turkish news.
• The dataset's last update was in March 2023, which may limit its recency for current events analysis.

Provenance

Source

Hugging Face, uploaded by user 'batubayk'.

Collection Method

Created for research on abstractive text summarization for agglutinative languages, as per the cited paper.

Freshness

Last updated on 2023-03-04.

Geography

Primarily Turkish language ('Language:tr'), with a 'Region:us' tag indicating potential US focus or source.